How do I change the name of the Default Bot or another Bot?
The Administrator of the account will need to be the one who changes the name for Bots. With the administrator role, you can then: 1. Head to Settings > Bots 2. Click on the Bot you want to rename 3. Type out the new name of the bot in the name field 4. You're all set!Can we get an extension on trial? How long will the trial last?
When you're successfully signed up for the eesel dashboard, you will get 7 days of trial as a business plan so you can explore the eesel product. After that time, the trial will automatically pause if you want to extend it further you can login again to your account and click "extend trial" and you will receive a further 7 days. \ **Note:** You can't extend the trial after 14 days.Why did the crawl get stuck?
Crawling issues usually come down to one of these reasons: * The URL is private, so our crawler can’t access it. * The link contains a large number of pages or heavy content, which can take longer to crawl. You can either wait for it to finish or try re-adding the links by refreshing the page. * The sitemap hasn’t been picked up properly. * The URL is for a help center or other platform and should be used by our help center crawler instead (see: integrations) If you are facing these issues, you can try a few ways by using our [Browser extension](https://chromewebstore.google.com/detail/eesel-ai-chatgpt-sidebar/ejhkkbilnpifailgngpkgmiofhioacjd?hl=en) to add specific pages by open the AI Copilot sidebar when you open the specific page and click "add page" (This way will work for private/authenticated sites as well). Re-add the sitemap directly to the crawler.How do I correct the bot's responses?
To provide specific instructions to correct your AI bot's responses, you can use our edit button. To do so: 1. Ask the AI a question in the "chat" tab of the dashboard. 2. Click the "edit" button at the bottom of the response you'd like to correct. 3. This will open up a new chat interface where you can have a conversation with the bot about what to change in its response. 4. You should explain to the bot what parts were bad, and what it should have said instead. 5. The bot will revise its response, and confirm whether you are happy with it. 6. Make sure you confirm you'd like that to be the updated way it response to similar queries. 7. The bot will update its Articles to incorporate this new training, and future responses will use this knowledge.What are Articles? How do I use them?
Articles are a quick, simple way of adding straightforward knowledge to your bot. They will be used as a source of information to answer questions. You can draft articles, edit them, and set them live to be used to answer queries. Articles are also what is generated when you have used the [past ticket training](#what-is-the-process-of-training-on-past-tickets-issues) feature, and the [corrections ](#how-do-i-correct-the-bots-responses)feature when you edit the AI's response. **How to create, edit, and use Articles** 1. To create an Article, simply head to the Article tab and click "New Article". 2. Ensure it contains a straight forward and clear title 3. Write out any content that you'd like the bot to know in plain text 4. Click "Add article" to create the Draft. 5. Select any articles you want to go live with, and hit "Set status to live" to enable the bot to use the Articles in its responses. At any point in time you can had back to the Articles tab, set the Articles back to draft or edit them live to ensure the bot has the most up to date information.How do I train on past tickets/issues?
**What is past ticket training?** Training your AI bot with past tickets allows eesel to pull selected tickets from your helpdesk and convert them into help articles. You can specify which tickets using a query for ticket parameters like a date range, closed, which agent handled them, specific tags, and more. The AI processes the selected tickets, and then articles are created based on them. These articles are then used to improve your bot's accuracy and provide more relevant, context-aware responses. The past ticket training will not include any attachments or internal notes by default. The articles will not be adding live instantly, nor are they added to your public help center. The articles will only appear inside the eesel dashboard for the bot to use. You can find them on the "Articles" tab on the left sidebar. Once the ticket training is completed, you can: * Review the articles that were generated * Edit or remove any that don't suit what you want the bot to know * Set them live for the bot to use for its responses **Who can use past ticket training?** Past ticket training is available for certain helpdesks users, who are also: * Trial users (with limited usage, and only for some helpdesks) * Business Plan customers * Custom Plan customers The following helpdesks have trial versions available for testing: * Zendesk * Gorgias * Freshdesk * Jira For other helpdesk inquiries, we may be able to conduct past ticket training upon request. Please reach out toHow do I get in touch with support?
If you've reviewed our help center and are still unable to find an answer to your question, then please reach out. To get in touch with support, please: 1. Head to your dashboard 2. Click Help and Support in the bottom left. 3. Click "Chat with us".What are the eesel support hours?
The eesel AI team usually operates between 9AM-5PM AEST as we're primarily based in Australia. However, some of our team will be online after 5PM on some days to help accommodate other time zones. We'll do our best to ensure we get to your questions as soon as we can! We aim to get back to you within 24 hours.What kind of support will I get for configuring our setup?
Each customer of ours will be able to get setup support for default integrations without extra cost. However, for more complex prompt engineering and actions, we do have limited availability. We can offer an Engineering Consultation Add-On for prompt engineering support and complex configuration. For more details about this, check out our [pricing page](https://www.eesel.ai/pricing) or [pricing help page](/legacy/pricing-admin-and-more/pricing).I want to delete my account and/or data.
You can delete your account by heading to your settings (top left bot dropdown menu) > Settings > General and click “Delete workspace”. Please note: you cannot undo this action. If you just want your data removed, you can remove specific links using the integration menu on the left hand side - or if you head to the "overview" page for integrations, you can uninstall specific integrations there also If you’re looking for custom data retention options, this is a custom item. Let me know what kind of custom data retention you’re looking for and we can explore a custom add-on once on a paid plan.I want to have a call/meeting with your team for a demo or to discuss other things.
You can find our demo meeting times here:Can I train the AI over more than 100 web pages or product pages in the trial?
Unfortunately, there is a crawl limitation on the trial (though the crawler by default does have a rather large 'leeway' range for larger sites). Once on a paid plan, we can crawl your entire website to ensure that bot has all the knowledge contained there.Can I sync my knowledge bases/website/help center?
Syncing is available upon request as part of a paid plan. You won’t be able to enable live syncing on a trial account. You can 'recrawl' your site manually using the website crawler as needed on the trial, then once on a paid plan please share how frequently you'd like it to be crawled and what bots and URLs you would like to apply this sync for.Can you exclude PII when training on past tickets?
It's possible for us to exclude PII when training on past tickets, upon request and only on a paid plan. Trial users who use the past ticket training tool will not be able to alter this.\ \ We recommend that in order to train the AI on past tickets accurately, it needs to be able to see the human agent replies, the original customer query and any relevant fields that directed the response, and how the ticket was resolved.Can I extend my trial?
Trial accounts have 1 week of trial time by default. If you need more time to test specific features, or other extenuating circumstances have prevented you from testing during that week, please reach out to the team to ask for an extension via the in-app messaging system or by emailingHow to change the avatar/logo/image of the chat bubble?
You can search for your company’s logo or picture on the internet, or upload an image to Imgur. Right click and select "copy image address". After you have the link, paste it into the Avatar link field in your chat bubble config page. Feel free to reach out toHow do I get the bot to read internal notes and use all data from a ticket?
You can use the "Read Context" tool/action to pull full information from ticket fields in order to use that as part of the response. See info about how to use this action [here.](/legacy/getting-started/prompt-and-actions#suggested-prompts) **What is this?** This tool allows your bot to access context information that was passed from the triggering system (like a helpdesk webhook). It provides access to metadata and additional information about the current conversation or ticket. **Common use cases:** 1. **full\_ticket** - Access complete ticket information including all comments with public/internal labels 2. **ticket\_id** - Get the current ticket ID 3. **current\_user\_email** - Access the email of the ticket requester **Why use this?** 1. Access rich context that may not be available in the conversation history 2. Distinguish between public customer comments and internal staff notes 3. Get structured data about the ticket for more informed responsesAre there user based permissions or access levels? How do I change the admin of an account?
Right now, the admin of an account is only differentiated from a regular member of an account by their ability to delete the workspace or rename bots. The majority of features in the eesel dash are available for all members to use as well. We can offer user access controls, on request, as a custom item on a paid plan. Please send through the details of what sort of UAC you'd prefer toDoes eesel AI support voice calls?
No, and it's not on our roadmap at the moment.How many users can I invite/add on each plan?
There are no limitations to the number of users or seats you can have. However, as all users can edit and customize bot prompts and knowledge, we generally recommend limiting it to those required for backend processes. Your agents will be able to test and review responses via the integrations you add the bots to.Is there a limit to the training or integrations added to a bot?
There are some limitations for trial accounts in terms of training volume for default integrations, but all default integrations have no strict limits on a paid account. For custom integrations, we will provide a quote based on volume.How do I change or customize the welcome message for the Zendesk Messaging Chat Widget?
1. Open your eesel AI dashboard. 2. Go to Zendesk Integration > AI Agents > AI Agents for Messaging. 3. Under Appearances, find the Welcome Message field. 4. Add or edit your custom greeting there.Is there an HTML/source view for the prompt editor?
No, there isn’t an HTML or source view for the prompt editor at the moment. We’re working on a feature that’ll let you copy prompts between bots while keeping the rich text formatting, but it’s not available just yet.Does eesel support ERP?
No, not out of the box. However, you can use our [Custom API add-on](/legacy/integrations/new-or-custom-integration/custom-actions-apis) to build integrations and actions across platforms we don't have integrations with and customise your agent's responsibilities and tasks.Help desks
Wikis
Internal team chats
| Slack | /pages/D0TwUOqeLfvAsv2FUdFA | /files/3KCdzhG2cc7PiDFkVf99 |
| Teams | /pages/h6S7Z80bjWjLwXFRijXr | /files/8rndYhcHeTSPQuz3yKg4 |
| Custom Integration | /pages/dDFpTNuqXuqN2OEjoHIB | /files/G7UKQOl3dnUD20Qc1Dsy |
More
Business Hours
The bot is aware of the time in UTC time zone. You can use the prompt to explain to the bot what your business's hours are, and what it should do if it's within or outside of those hours. Make sure you add the **"think" action** through the 'create new action' button, and then mention the "think" action in the prompt. You also need to include your time zone based around UTC. An example prompt might be:\ \ `"Business hours context to ALWAYS follow: our business operates on Monday - Friday. The time zone we operate in is UTC-4. Always use the think tool to think about the current time and our business hours before making a decision on whether its in business hours."`Clarifying Questions
Your bot is able to ask clarifying questions that enable the AI to ensure it's addressing the right questions with the right information. An easy way to use the prompt to ensure it is asking clarifying questions is to include something like: "If a user asks a question, always ensure you clarify with them what product they are referring to (if they haven't already specified)." Or, you can add in more specific guidelines for clarifying questions: `"At the start of every conversation, always ask the user to provide the following:` * `Email address` * `Product name` * `A summary of their issue` `Once this has been provided, then you can assist with their query."`Language
By default, eesel AI has access to 80+ languages and is able to respond and understand all of these languages. Using the prompt, you can tell the bot to either: `"Always respond in the language of the user."` Or `"Only ever respond in English".` If you'd like to provide a translation of the messages, like as an internal note or to the user, you can then specify that the bot should answer the query, and also provide a translated version etc.What to do when the AI doesn't know the answer
Having guardrails for when the bot doesn't know the answer is always recommended to include. Depending on how you want the AI to handle things in this instance, there are a few options: * Ask clarifying questions (see above) to ensure the AI understands what is being asked * Let the user know that it's unable to answer their query and they should contact another support channel * Let the user know that it's unable to answer their query, and that they will escalate them to an agent to handle * Let the user know that it's unable to answer their query, and if they would like to create a support ticket. In each instance, these can be used in conjunction with the suitable actions that align with the integrations you have. For example, using the `zendesk_create_ticket` action or the `freshdesk_assign_ticket` action.Escalation
Escalation is important for instances where you'd like the user to be able to access a chat with a live human. In more instances, this is highly recommended to avoid frustration with customers who prefer to speak to humans. Within the prompt, you can add in a simple `"If the user asks to speak to a human, then always use 'zendesk_assign_ticket' action with new_assignee='[Sales Team Assignee ID]'"` (or the equivalent) helpdesk action. This will allow you to assign tickets to an agent or a team, based on the ID of that agent or team.When you don't want the bot to reply
In certain circumstances, you may not want the bot to reply. This might be useful if: * You don't want the bot to reply to spam, or harmful messages * You don't want the bot to attempt to handle sensitive tickets, so that it can hand off immediately To do this, you can use the **"do not reply" action.** Simply add the action via "create new action" and add to the prompt something like: `"If the user's question is about purchasing a custom plan, then use 'do_not_reply'. Then, use 'zendesk_internal_note' to leave a note about why you didn't leave a reply, and what the suggested response might be."`Think
**Think** action: add the think action to enable the bot to spend more time reasoning, useful for complex conditionals. Example: "Use the '`think`' action when you decide which set of instructions you follow below, based on the product the user has specified."Do not reply
**Do not reply** action: add this action to tell the bot in which instances it should not reply at all. Example: "If the query is using rude words, is confrontational, and extremely negative, then use the '`Do not reply`' action."Read full document
**Read full document** action: add this action if you want eesel to read an entire document in certain scenarios to ensure it's able to pull full context from important documents. Example: Always use the `retrieve_full_document` tool to find the most relevant product document and read it fully, before responding to the query.Use knowledge from another bot
**Search documents of another bot** 1. This is used if you want a new bot but want it to be able to access the training resources/knowledge of another bot. 2. Example prompt: "If the question is to do with "Product A", then always use `retrieve_sibling_documents` with "2da11eac-5585-4837-91b9-09b97f631288" to retrieve the relevant documents for "Product A" before answering"Consult/direct query to another bot
**Consult another bot** 1. This is used if you want a new bot but you want it to be able to access the training and *actions/customization* of another bot. Useful in instances where you want one bot to welcome/triage the queries, and then consult another bot to handle it entirely.This is useful when: 1. Situations when adding all Sources one bot alone leads to inaccurate answers. 2. Helpdesk triaging needs to be set up 3. You want to ask several clarifying questions before routing 2. Example prompt: "If the question is about "Product A", then always use consult\_bot with bot ID:123489 to answer the question." 3. The bot ID can be found by going to the chat page of the dashboard of the target bot, looking at the URL of the bot and copying the string of letters and number found between /v2/ and /chat.Generate suggestions
Use this action to allow your bot to generate contextual quick suggestions (chips/buttons) to guide the user to next best steps. Example prompt: If someone asks about looking for help with a product, always use ui\_generate\_suggestions after answering, to generate relevant questions that might help better direct the user's query.Read context
`read_context` * **Purpose:** Reads context information from the current ticket invocation, including the full ticket details with internal notes. * **Key Params:** `key: str` (the context key to retrieve, e.g., "full\_ticket", "ticket\_id", "current\_user\_email") * **When/How to Use:** Use this tool to access detailed ticket information that isn't available in the standard conversation context. The most useful key is "full\_ticket" which provides the complete ticket data including subject, description, and all comments (both public and internal notes). This is particularly valuable when you need to understand the full context of a ticket, including internal notes that agents have left, or when you need specific ticket metadata. * **Example Prompt Instruction:** > If you need to understand the complete ticket history including internal notes from agents, or if you need specific ticket details beyond what's in the conversation, call read\_context with key='full\_ticket'. This will give you access to the complete ticket structure including all internal communications. > > To get the current ticket ID, call read\_context with key='ticket\_id'. > > To get the current user's email address, call read\_context with key='current\_user\_email'. **Available Context Keys:** * `full_ticket`: Complete ticket data including subject, description, and all comments (public and internal) * `ticket_id`: The current ticket's ID * `current_user_email`: Email address of the ticket requesterCan I adjust the bot's tone based on the channel?
Yes! The way to do this depends on what platform the bot is using to respond. For helpdesks, you can use one of the helpdesk actions to pull context about the 'channel'. (Like the 'get field values' action or the ['read context' action](#read-context)), and then adding to the prompt to explain to the bot what kind of tone or formatting it should use based on what that action returns. For other platforms, it's recommended to create unique bots for different channels or audience segments. That way you can either connect a bot completely uniquely prompted for that channel's tone, or use the '[consult bot' action](#consult-direct-query-to-another-bot) to route conversations to the right bot that is prompted to handle things differently, based on the context of the query.Can we set up a delay for responses to look more human?
At the moment, we're working on a delay for responses.Can the AI Agent leave a public reply and an internal note/comment at the same time?
Yes, this is possible. The bot can be set up so that one action leaves an internal note, while the AI Agent sends a public reply to the customer. You need to set the bot to leave public replies as a default behaviour and then add the 'leave internal note' action to the prompt in the Customize tab. Then, explain in the prompt that the AI should always leave a note with "X" information each time it receives a ticket. For example, it will reply to the customer with an answer and leave a note with a translated version of the ticket.Can eesel AI chat bubble be added to a mobile app?
This is not possible natively, but we can arrange for this as a custom item.Close ticket
`zendesk_close_ticket` * **Purpose:** Close a Zendesk ticket. * **Key Params:** `close: true` * **When/How to Use:** Use this tool when the user explicitly indicates their issue is resolved, they are satisfied with the assistance, or the conversation reaches a natural conclusion where closure is appropriate according to your support process. The ticket ID for the current conversation is automatically available to the tool. * **Example Prompt Instruction:** > If the user explicitly states their problem is solved, they are satisfied with the outcome, or the conversation has concluded and no further action is required, you must confirm this understanding with the user. If they agree, call the zendesk\_close\_ticket tool to mark the ticket as closed.Leave internal note
`zendesk_leave_internal_note` * **Purpose:** Add a private internal note to a Zendesk ticket. * **Key Params:** `note: str` * **When/How to Use:** Use this tool to add context, summarize the AI's interaction, or flag important information for a human agent who might review the ticket later. This note is not visible to the end-user. Use this when the AI performs an action (like escalating) or when the conversation contains nuances a human agent should be aware of. The ticket ID is automatically available. * **Example Prompt Instruction:** > If you escalate a conversation to a human agent, or if you are unable to fully resolve the user's query but have provided some assistance, you must call zendesk\_leave\_internal\_note. The note parameter should be a brief summary of the user's request and the steps you took or information you provided before the handover.Tag ticket
`zendesk_tag_ticket` * **Purpose:** Add tags to a Zendesk ticket for categorization and workflow automation. * **Key Params:** `tag: str` (space-separated) * **When/How to Use:** Use this tool when the conversation's topic or resolution falls into specific categories you use for reporting, routing, or automation. Instruct the AI to identify keywords or themes in the conversation to determine the appropriate tags. If applying multiple tags in one call, list them space-separated in the `tag` parameter string. The ticket ID is automatically available. * **Example Prompt Instruction:** > After assisting the user, analyze the conversation's main topic and the resolution. If the user's request was related to 'billing', 'shipping', or 'product support', you must call zendesk\_tag\_ticket. The tag parameter should be the relevant tag (e.g., 'billing', 'shipping', 'product-support'). If the issue was successfully resolved by you, also add the tag 'ai-resolved'. If both apply, use space-separated tags like 'shipping ai-resolved'.Assign ticket (escalation)
`zendesk_assign_ticket` * **Key Params:** `new_assignee: str` * **When/How to Use:** Use this tool when the query requires human expertise that the AI cannot provide, or when the conversation needs to be routed to a specific team (represented by an assignee). You will need to define how the AI identifies the correct `new_assignee` identifier (which is typically an ID or a precisely matched name depending on your Zendesk setup and the tool's exact implementation, but is not an email address). The ticket ID is automatically available. * **Example Prompt Instruction:** > If the user's request involves a complex technical issue you cannot resolve, or if the user specifically asks to speak with a human agent, you must inform the user you are escalating the issue and then call zendesk\_assign\_ticket. Set the new\_assignee parameter to the specific ID or identifier for your 'Technical Support Team' group.Chat handover (Zendesk messenger)
`zendesk_chat_handover` * **Purpose:** Handover a Zendesk Messenger chat to a human agent. * **Key Params:** — (No explicit parameters) * **When/How to Use:** This is specifically for chat interactions. Use this when the AI determines it cannot resolve the user's query in a chat format and human intervention is required immediately. * **Example Prompt Instruction:** > In a chat conversation, if the user's query is unclear, requires account-specific actions you cannot perform, or the user explicitly asks to chat with a human, you must inform the user that you are connecting them with an agent and then call the zendesk\_chat\_handover tool.Update ticket
`zendesk_update_ticket` * **Purpose:** Updates a Zendesk ticket with the provided payload. * **Key Params:** `payload: str` (JSON string) * **When/How to Use:** Use this for modifying various ticket fields (status, priority, custom fields, etc.). The ticket ID for the current conversation is automatically available. You must instruct the AI to construct the JSON payload based on the desired updates indicated by the user or the workflow logic. * **Example Prompt Instruction:** > If the user indicates their issue is resolved and you have confirmed it, or if the workflow requires setting the ticket status to 'solved', you must call zendesk\_update\_ticket. Set the payload parameter to a JSON string that updates the status, e.g., payload='{"status": "solved"}'. > > If the user reports a high-priority issue you are escalating, you might call `zendesk_update_ticket` with `payload='{"priority": "high"}'`. > > To update a custom field (ID 123), call `zendesk_update_ticket` with `payload='{"custom_fields": [{"id": 123, "value": "new value"}]}'`.Get ticket fields
`zendesk_get_ticket_fields` * **Purpose:** Retrieves specified fields from a Zendesk ticket. * **Key Params:** `fields: str` (comma-separated) * **When/How to Use:** Use this when you need specific information from the current ticket (e.g., its current status, assignee, or a custom field value) to inform your next action or response. You must specify the field names (standard or custom field IDs as strings) you need. The ticket ID is automatically available. * **Example Prompt Instruction:** > Before escalating a ticket, you should check its current status and assignee. Call zendesk\_get\_ticket\_fields with the parameter fields='status,assignee\_id'. Use the results to inform the user about the current state or to decide who to assign it to next. > > If a user's query depends on a custom field (ID 456), call `zendesk_get_ticket_fields` with `fields='456'`.Read Context
`read_context` * **Purpose:** Reads context information from the current ticket invocation, including the full ticket details with internal notes. * **Key Params:** `key: str` (the context key to retrieve, e.g., "full\_ticket", "ticket\_id", "current\_user\_email") * **When/How to Use:** Use this tool to access detailed ticket information that isn't available in the standard conversation context. The most useful key is "full\_ticket" which provides the complete ticket data including subject, description, and all comments (both public and internal notes). This is particularly valuable when you need to understand the full context of a ticket, including internal notes that agents have left, or when you need specific ticket metadata. * **Example Prompt Instruction:** > If you need to understand the complete ticket history including internal notes from agents, or if you need specific ticket details beyond what's in the conversation, call read\_context with key='full\_ticket'. This will give you access to the complete ticket structure including all internal communications. > > To get the current ticket ID, call read\_context with key='ticket\_id'. > > To get the current user's email address, call read\_context with key='current\_user\_email'. **Available Context Keys:** * `full_ticket`: Complete ticket data including subject, description, and all comments (public and internal) * `ticket_id`: The current ticket's ID * `current_user_email`: Email address of the ticket requesterCreate ticket
`zendesk_create_ticket` * **Purpose:** Creates a new ticket in Zendesk. * **Key Params:** `summary: str, name: str, email: str`, optional `tags: List[str]`, optional `group_id: int`, optional `assignee_id: int` * **When/How to Use:** Use when a user's request warrants creating a new, separate ticket, such as a new inquiry identified within an existing conversation, or a request received via a non-ticket channel that needs formal tracking. You must gather the `summary`, `name`, and `email` of the requester. Optional parameters like `tags`, `group_id`, or `assignee_id` can be set based on workflow rules or user input. * **Example Prompt Instruction:** > If a user asks a new question that is unrelated to the current conversation's topic, offer to create a new ticket for that specific question. If they agree, collect a brief summary of the new question, the user's name, and email address. Call zendesk\_create\_ticket with the collected summary, name (from user context), and email (from user context). Add a tag like tags=\['new\_inquiry'] or specify an group\_id if needed.How do I set up the AI as a unique user in my Zendesk?
If you need to change the user that eesel AI uses to send replies and updates inside Zendesk, follow these steps: **Step 1: Create a New User in Zendesk** 1. Log into your Zendesk account. 2. Navigate to Admin Center > People > Agents 3. Click Add user and enter the necessary details (name, email, role). 4. Save the user. **Step 2: Find the User ID** The User ID is a unique identifier required for eesel AI. To find it: 1. In Zendesk, go to Admin Center > People > Search Users and find the newly created user. 2. Click on the user’s name to open their profile. 3. Look at the URL in your browser’s address bar—it will be in the format: ``` https://yourcompany.zendesk.com/agent/users/123456789 ``` 4. The number at the end (e.g., `123456789`) is the User ID. Copy this. **Step 3: Update eesel AI’s Response Settings** 1. Log into eesel AI. 2. In the sidebar, navigate to Zendesk > AI Agent. 3. Find the field titled "What user should the bot respond as?" 4. Paste the copied User ID into this field. 5. Save the changes. eesel AI will now send replies and updates as the selected user inside Zendesk. You can also use Step 2 to find an existing user you want the bot to reply as, rather than the new ID.Can eesel trigger a side conversation in Zendesk?
No, eesel does not interact with side conversations inside Zendesk.Can eesel detect in real-time when there is an agent available for escalation?
No, eesel is not able to determine whether there is a live agent available.
Create ticket
`freshdesk_create_ticket` * **Purpose:** Creates a new Freshdesk ticket. (Note: This definition appears to be the same as the previous one). * **Key Params:** `subject, description, email, priority, status` * **When/How to Use:** Use when a user reports a bug, requests a new feature, or has a complex issue that requires tracking as a formal ticket rather than being resolved immediately in the current conversation. You will need to instruct the AI to gather the required parameters from the user. The current conversation ID is automatically available for context if needed, but the tool creates a *new* ticket. * **Example Prompt Instruction:** > If the user is reporting a potential bug or requesting a new feature, you must offer to create a support ticket for them in our Freshdesk system. If they accept, collect their name, email address, a brief summary of the issue (for the subject), and a detailed description of the problem or request. Once you have this information, call freshdesk\_create\_ticket with the gathered subject, description, and email. Set priority to 'Low' and status to 'Open' by default, unless the user indicates high urgency.Tag ticket
`freshdesk_tag_ticket` * **Purpose:** Add tags to a Freshdesk ticket. * **Key Params:** `tags: List[str]` * **When/How to Use:** Similar to other tagging Actions for categorization and workflow. Note this tool takes a `List[str]` for the `tags` parameter, meaning you should provide tags in a list format like `['tag1', 'tag2']`. The ticket ID for the current conversation (if applicable) is automatically available. If tagging a ticket just created by `freshdesk_create_ticket`, the ID might need to be handled via workflow logic or be available from the create tool's response. * **Example Prompt Instruction:** > After successfully creating a Freshdesk ticket using freshdesk\_create\_ticket, you must also call freshdesk\_tag\_ticket using the ticket ID of the newly created ticket (ensure this ID is available after the create call). Add the tag 'created\_by\_ai' and any other relevant tags based on the user's request, such as 'bug' or 'feature\_request'. The tags parameter must be a list of strings, for example: tags=\['created\_by\_ai', 'bug'].Close ticket
`freshdesk_close_ticket` * **Purpose:** Close a Freshdesk ticket. * **Key Params:** `close: true` * **When/How to Use:** Use this tool when the user explicitly indicates their issue is resolved, they are satisfied with the assistance, or the conversation reaches a natural conclusion where closure is appropriate according to your support process. The ticket ID for the current conversation is automatically available to the tool. * **Example Prompt Instruction:** > If the user explicitly states their problem is solved, they are satisfied with the outcome, or the conversation has concluded and no further action is required, you must confirm this understanding with the user. If they agree, call the Freshdesk\_close\_ticket tool to mark the ticket as closed.Leave internal note
`freshdesk_leave_internal_note` * **Purpose:** Leaves an internal note on the ticket. * **Key Params:** `note: str`, optional `notify_emails: List[str]` * **When/How to Use:** Use this to add private notes to the current Freshdesk ticket for internal team members. You can optionally specify a list of agent emails to notify about the note. The ticket ID is automatically available. * **Example Prompt Instruction:** > If you escalate a conversation to a human agent, or if you need to summarize the AI's actions for the next agent, call freshdesk\_leave\_internal\_note. The note parameter should contain the summary. For urgent escalations that need specific attention, you can also include notify\_emails=\['Assign ticket (escalation)
`freshdesk_assign_ticket` * **Purpose:** Assigns a Freshdesk ticket to a specific agent. * **Key Params:** `assignee_id: int` * **When/How to Use:** Use this tool to route the current ticket to a specific agent based on their ID. The ticket ID is automatically available. Define rules for which agent ID to use based on the conversation topic or user request. * **Example Prompt Instruction:** > If the user's query requires the attention of an agent specializing in integrations, inform the user you are assigning the ticket to the integrations specialist and call freshdesk\_assign\_ticket. Set the assignee\_id parameter to the integer ID of the integrations agent, e.g., assignee\_id=12345.Assign ticket to group
`freshdesk_assign_ticket_to_group` * **Purpose:** Assigns a Freshdesk ticket to a specific group. * **Key Params:** `group_id: int` * **When/How to Use:** Use this tool to route the current ticket to a specific support group or team based on their group ID. The ticket ID is automatically available from the conversation context. This is particularly useful for routing tickets to specialized teams (e.g., technical support, billing, escalations) based on the nature of the customer inquiry. * **Example Prompt Instruction:** > If the user's issue requires specialized handling (e.g., technical issues go to the technical support group, billing issues to the billing team), inform the user you are transferring their ticket to the appropriate team and call freshdesk\_assign\_ticket\_to\_group. Set the group\_id parameter to the appropriate group's integer ID. For example: for technical issues use group\_id=67890, for billing use group\_id=54321. Always confirm with the user before transferring their ticket to a different group.Update ticket fields
* `freshdesk_update_ticket` * **Purpose:** Updates a Freshdesk ticket with the provided payload. * **Key Params:** `payload: str` (JSON string) * **When/How to Use:** Use this for modifying various ticket fields (status, priority, custom fields, etc.). The ticket ID for the current conversation is automatically available. You must instruct the AI to construct the JSON payload based on the desired updates indicated by the user or the workflow logic. * **Example Prompt Instruction:** > If the user indicates their issue is resolved and you have confirmed it, or if the workflow requires setting the ticket status to 'solved', you must call freshdesk\_update\_ticket. Set the payload parameter to a JSON string that updates the status, e.g., payload='{"status": "solved"}'. > > If the user reports a high-priority issue you are escalating, you might call `freshdesk_update_ticket` with `payload='{"priority": "high"}'`. > > To update a custom field (ID 123), call `freshdesk_update_ticket` with `payload='{"custom_fields": [{"id": 123, "value": "new value"}]}'`.Tag conversation
`intercom_tag_conversation` * **Purpose:** Add tags to an Intercom conversation. * **Key Params:** `tag: str` * **When/How to Use:** Similar to `zendesk_tag_ticket`. Use for categorization, reporting, or triggering workflows based on conversation topics. Note that this tool takes a single string `tag`. The conversation ID is automatically available. * **Example Prompt Instruction:** > Upon successfully answering a user's question about feature usage, you must call intercom\_tag\_conversation. Set the tag parameter to 'feature-inquiry'. If the user was asking about a known issue, make a separate call to intercom\_tag\_conversation with the tag 'known-issue'.Close conversation
`intercom_close_conversation` * **Purpose:** Close an Intercom thread after a short delay (\~30 seconds). * **Key Params:** — (No explicit parameters) * **When/How to Use:** Similar to `zendesk_close_ticket`. Use when the conversation is resolved and no further action is needed. The approximately 30-second delay is built-in. The conversation ID is automatically available. * **Example Prompt Instruction:** > If the user confirms their issue is resolved and expresses satisfaction, or if the conversation has been inactive for several minutes after you provided a solution, you must call intercom\_close\_conversation to close the thread after the standard delay.Leave internal note
`intercom_leave_internal_note` * **Purpose:** Leaves an internal note on the ticket. * **Key Params:** `note: str` * **When/How to Use:** Use this action to leave internal notes on conversations in Intercom. Describe what you want the internal note to be - this could be a draft reply for your agents to pick up, a translation of the user message and so on. * **Example Prompt Instruction:** > If you escalate a conversation to a human agent, or if you need to summarize the AI's actions for the next agent, call intercom\_leave\_internal\_note. The note parameter should contain the summary. For urgent escalations that need specific attention, you can also include notify\_emails=\['Assign conversation (escalation)
`intercom_assign_conversation` * **Purpose:** Assign an Intercom conversation to a specific team or admin. * **Key Params:** `assignee_id: str` * **When/How to Use:** Similar to `zendesk_assign_ticket`. Use this for routing conversations that require human follow-up or expertise from a specific team. Define how the AI determines the appropriate `assignee_id`. This parameter requires the Intercom ID for the admin or team, not an email address. The conversation ID is automatically available. * **Example Prompt Instruction:** > If the user asks a question about their billing history or requires an action related to their subscription, you must inform them you are escalating their request to the billing team and call intercom\_assign\_conversation. Set the assignee\_id parameter to the specific Intercom ID for your 'Billing Team' inbox or a specific billing specialist admin.Get ticket fields
`Intercom_get_ticket_fields` * **Purpose:** Retrieves specified fields from an Intercom ticket. * **Key Params:** `fields: str` (comma-separated) * **When/How to Use:** Use this when you need specific information from the current ticket (e.g., its current status, assignee, or a custom field value) to inform your next action or response. You must specify the field names (standard or custom field IDs as strings) you need. The ticket ID is automatically available. * **Example Prompt Instruction:** > Before escalating a ticket, you should check its current status and assignee. Call intercom\_get\_ticket\_fields with the parameter fields='status,assignee\_id'. Use the results to inform the user about the current state or to decide who to assign it to next. > > If a user's query depends on a custom field (ID 456), call `intercom_get_ticket_fields` with `fields='456'`.Get conversation fields
`Intercom_get_conversation_fields` * **Purpose:** Retrieves specified fields from an Intercom ticket. * **Key Params:** `fields: str` (comma-separated) * **When/How to Use:** Use this when you need specific information from the current conversation (e.g., its current status, assignee, or a custom field value) to inform your next action or response. You must specify the field names (standard or custom field IDs as strings) you need. The conversation ID is automatically available. * **Example Prompt Instruction:** > Before escalating a conversation, you should check its current status and assignee. Call intercom\_get\_conversation\_fields with the parameter fields='status,assignee\_id'. Use the results to inform the user about the current state or to decide who to assign it to next. > > If a user's query depends on a custom field (ID 456), call `intercom_get_conversation_fields` with `fields='456'`.Un-assign conversation
`Intercom_unassign_conversation` * **Purpose:** Use this action to unassign conversations in Intercom. * **Key Params:** * **When/How to Use:** Simply describe when you want the conversation unassigned, and the AI will handle this. This is useful for removing an assignee from a conversation, making it available for others to pick up. * **Example Prompt Instruction:** >User lookup
`Intercom_get_conversation_fields` * **Purpose:** Allow your bot to look up a Intercom user using their email address. * **Key Params:** * **When/How to Use:** * **Example Prompt Instruction:** >Update conversation
`Intercom_update_conversation` * **Purpose:** Use this action to update a conversation in Intercom. * **Key Params:** * **When/How to Use:** Define the conditions and details you want to change (such as status, priority, or assignee), and the AI will automatically update the conversation accordingly. * **Example Prompt Instruction:** > If the user indicates their issue is resolved and you have confirmed it, or if the workflow requires setting the ticket status to 'solved', you must call intercom\_update\_conversation. Set the payload parameter to a JSON string that updates the status, e.g., payload='{"status": "solved"}'. > > If the user reports a high-priority issue you are escalating, you might call `intercom_update_ticket` with `payload='{"priority": "high"}'`. > > To update a custom field (ID 123), call `intercom_update_ticket` with `payload='{"custom_fields": [{"id": 123, "value": "new value"}]}'`.Does eesel support Intercom back-office tickets?
No.Tag conversation
`gorgias_tag_ticket` * **Purpose:** Applies tags to a ticket in the Gorgias helpdesk system. * **Key Params:** `tags: List[str]` * **When/How to Use:** Use this for categorizing Gorgias tickets based on conversation topics or resolution types. You must provide the tags as a list of strings. The ticket ID for the current conversation is automatically available. * **Example Prompt Instruction:** > Upon successfully resolving a user's issue related to shipping, call gorgias\_tag\_ticket with the parameter tags=\['shipping', 'resolved-by-ai']. If the user was asking about a specific product, also include a product tag, e.g., tags=\['shipping', 'resolved-by-ai', 'product-X'].Create ticket
`gorgias_create_ticket` * **Purpose:** Creates a new ticket in Gorgias. * **Key Params:** `subject: str, description: str, email: str, to: str`, optional `priority: str`, optional `status: str`, optional `custom_fields: dict`, optional `channel: str`, optional `via: str`, optional `from_agent: bool` * **When/How to Use:** Use this when a user's request requires creating a brand new ticket in Gorgias, possibly for a new inquiry identified in a chat or social media interaction that needs formal tracking. You must gather the `subject`, `description`, and `email` of the requester, and specify the `to` email address of the Gorgias mailbox where the ticket should be created. Other parameters like priority or custom fields can be set based on context. * **Example Prompt Instruction:** > If a user submits feedback via a channel that doesn't automatically create a ticket (like an in-app message you handle conversationally), and the feedback needs formal tracking, offer to create a ticket. If accepted, collect the feedback subject, detailed description, and user's email. Call gorgias\_create\_ticket with the collected info, setting to='Leave internal note
`gorgias_leave_internal_note` * **Purpose:** Leaves an internal note on the ticket. * **Key Params:** `note: str` * **When/How to Use:** Use this action to leave internal notes on conversations in Gorgias. Describe what you want the internal note to be - this could be a draft reply for your agents to pick up, a translation of the user message and so on. * **Example Prompt Instruction:** > If you escalate a conversation to a human agent, or if you need to summarize the AI's actions for the next agent, call gorgias\_leave\_internal\_note. The note parameter should contain the summary. For urgent escalations that need specific attention, you can also include notify\_emails=\['Assign ticket (escalation)
`gorgias_assign_ticket` * **Purpose:** Assigns a ticket to a specific user in the Gorgias helpdesk system. * **Key Params:** `assignee_id: int` * **When/How to Use:** Use this to route the current Gorgias ticket to a specific agent or team member identified by their user ID. The ticket ID is automatically available. * **Example Prompt Instruction:** > If the user's query requires a human agent for a complex return request, inform them you are escalating to the returns specialist and call gorgias\_assign\_ticket. Set the assignee\_id parameter to the integer ID of the returns specialist, e.g., assignee\_id=67890.Create issue
`jira_create_issue` * **Purpose:** File a Jira issue. * **Key Params:** `project_key, summary, description` * **When/How to Use:** Primarily used for reporting bugs or suggesting improvements that need engineering or product team attention tracked in Jira. You must instruct the AI to identify these scenarios and gather the necessary information. Define the default `project_key` to use. * **Example Prompt Instruction:** > If you are unable to help the user with their enquiry OR it appears that they are reporting a bug or suggesting a product improvement, you must offer to create an issue for the user in our engineering tracking system (Jira). If they accept, collect a brief summary of the issue and a more detailed description. Then, call jira\_create\_issue with project\_key set to 'ENG', using the collected information for the summary and description parameters.Label issue
`jira_label_issue` * **Purpose:** Adds labels to a Jira issue. * **Key Params:** `issue_key: str, labels: List[str]` * **When/How to Use:** Use this after a Jira issue has been created or identified (e.g., using `jira_create_issue` or `jira_search_issue`) to add relevant labels for categorization and filtering within Jira. You need the issue key and the list of labels. * **Example Prompt Instruction:** > After successfully creating a Jira issue using jira\_create\_issue (and obtaining the issue key from the tool's response), analyze the user's report for key topics (e.g., 'login', 'performance'). Call jira\_label\_issue using the new issue\_key and a list of relevant labels, e.g., labels=\['bug', 'reported-by-customer', 'severity-medium'].Search issue
`jira_search_issue` * **Purpose:** Searches for Jira issues using JQL. * **Key Params:** `jql: str` * **When/How to Use:** Use this when a user asks about the status of a known bug, feature request, or project item that is tracked in Jira. You must formulate a JQL query string based on the user's keywords or the issue they are referencing. * **Example Prompt Instruction:** > If the user asks about the status of a bug they previously reported, or asks if a known issue is fixed (and mentions keywords related to it), formulate a JQL query to search for the issue in Jira. Call jira\_search\_issue with the parameter jql='summary \~ "\[Keywords from user query]" AND project = ENG ORDER BY created DESC'. Present the search results to the user, focusing on the issue key, summary, and status.Assign issue (escalation)
`jira_assign_issue` * **Purpose:** Assigns an existing issue in Jira to a specific user. * **Key Params:** `issue_key: str, assignee_account_id: str` * **When/How to Use:** Use this to assign an existing Jira issue (identified by its key) to a specific team member or group within Jira, using their Jira account ID. This is useful for routing issues found via search or created by the AI to the correct owner. * **Example Prompt Instruction:** > If the user is reporting a front-end issue and you have found or created the relevant Jira ticket (issue key available), you should assign it to the front-end lead. Call jira\_assign\_issue with the issue\_key and the parameter assignee\_account\_id='\[Front-end Lead Jira Account ID]'.Comment on issue
`jira_comment_on_issue` * **Purpose:** Adds a comment to an existing issue in Jira. * **Key Params:** `issue_key: str, comment: str` * **When/How to Use:** Use this to add context from the customer conversation directly to the related Jira issue. This is useful for providing engineers with additional details, clarifying user impact, or adding follow-up information. You need the issue key and the content of the comment. * **Example Prompt Instruction:** > If a user provides important additional details or context about a bug or feature request after the Jira issue has been created, call jira\_comment\_on\_issue using the relevant issue\_key. Set the comment parameter to summarize the new information provided by the user.Connect Jira to eesel AI via actions
To connect your Jira account to eesel , you’ll need to generate an API token and provide a few details. Follow the steps below: #### 1. Generate an API Token Visit [this link](https://id.atlassian.com/manage-profile/security/api-tokens) to generate a new API token in your Atlassian account. #### 2. Create and Copy the Token Click "Create API Token", give your token a recognizable name, and click "Create".\ Then, copy the generated token to your clipboard — you’ll need it in a later step. #### 3. Find Your Jira Subdomain Locate your Jira site URL. It will look something like `yourcompany.atlassian.net`. Where `yourcompany` is your subdomain. #### 4. Enter Your Jira Email Use the email address associated with your Jira account. #### 5. Paste the API Token Paste the API token you copied earlier into the relevant field. #### 6. Connect Click "Connect Jira" to finalize the setup. #### Want a Full AI Agent Embedded in Jira? If you'd like a full AI agent embedded directly inside your Jira or JSM project, install the [AI for Jira Cloud](https://marketplace.atlassian.com/apps/1232959/ai-for-jira-cloud) app from the Atlassian Marketplace:Can I set a bot up to read and summarize specific Jira's based on their ID while working in Zendesk?
Yes, you can add the ‘search issues’ action found in the Jira category of the actions menu (see: customize tab > create new action). \ You can enable that action and then use the prompt to describe when the bot should use it inside Zendesk, or another platform. Check [this page](https://docs.eesel.ai/integrations/jira-service-management#search-issue) out for more instructions.I've installed the Jira app but eesel isn't operating in my Jira.
First, ensure that you have switched the AI Agent to either leave internal notes, or leave public replies, to see it responding to issues in Jira directly. Then, make sure you add a JSM Project Key to the Project Key field at the bottom of the AI Agent page. You can find the Project Key by looking at the acrnoym of letters at the start of each of your issues. For example: ITSM-913 is issue 913 of the ITSM project. Add in ITSM to the project key field.Can eesel connect with Jira Data Center?
No, eesel cannot connect to Jira Data Center.Change Ticket Priority
`hubspot_change_ticket_priority` * **Purpose:** Use this action to change the priority of a ticket in HubSpot. Simply describe when you want the ticket priority changed, and the AI will handle this. * **Key Params:** `priority, ticket_id` * **When/How to Use:** Use when a customer’s issue needs escalation or de-escalation based on urgency. Valid priority values are “LOW”, “MEDIUM”, “HIGH”, or “URGENT”. You must collect the ticket ID and determine the appropriate priority level based on the issue severity. * **Example Prompt Instruction:** > *If a customer reports a critical issue that affects their business operations or mentions urgency, offer to escalate their ticket priority. If they agree, collect the ticket ID and set the priority to “HIGH” or “URGENT” depending on the severity. For routine inquiries, you can suggest lowering priority to “LOW” or “MEDIUM”.*Add Note to Ticket
`hubspot_add_note_to_ticket` * **Purpose:** Adds an internal note to a HubSpot ticket that’s only visible to agents * **Key Params:** `ticket_id, note` * **When/How to Use:** Use for internal documentation, tracking investigation steps, or leaving notes for other agents. These notes are not visible to customers and help maintain context for future interactions. * **Example Prompt Instruction:** > *When you need to document internal information about a customer’s case, investigation steps taken, or leave instructions for other agents, use this tool to add internal notes. Always collect the ticket ID and write clear, informative notes that will help other agents understand the context.*Update ticket
`hubspot_update_ticket` * **Purpose:** Updates a HubSpot ticket’s title and description * **Key Params:** `title, description, ticket_id` * **When/How to Use:** Use when you need to modify ticket information to better reflect the customer’s issue or add clarification. Particularly useful when the original ticket title/description is unclear or incomplete. * **Example Prompt Instruction:** > *If a customer provides additional context about their issue or if the original ticket description is unclear, offer to update the ticket with more accurate information. Collect the ticket ID and ask for the updated title and description that better represents their issue.*Add comment to ticket
`hubspot_add_comment_to_ticket` * **Purpose:** Adds a public comment to a HubSpot ticket that customers can see * **Key Params**: `ticket_id, comment` * **When/How to Use:** Use to communicate directly with customers through the ticket system, provide updates, ask for additional information, or respond to their queries within the ticket thread. * **Example Prompt Instruction:** > *When you need to communicate with a customer through their ticket or provide an update on their case, use this tool to add a comment. The customer will be able to see this comment. Collect the ticket ID and write a professional, helpful comment addressing their needs or providing the requested information.*Can I change the ID the bot uses to respond?
Yes, you can add an ID in the field at the bottom of the AI Agents page in the eesel dashboard for HubSpot.Does eesel AI train on HubSpot deals and notes?
No, eesel is not currently able to learn from HubSpot deals and notes.What HubSpot services does eesel connect with?
eesel currently integrates with HubSpot’s CRM Conversations Inbox and Service Hub. It does not natively integrate with HubSpot Marketing features, so the exact compatibility depends on what data or workflows you’re looking to connect from HubSpot Marketing.How do I enable auto-sync for my website pages?
To enable auto-sync for your public or private website, you must be on a paid plan. Once on a plan, please reach out to support in-app or via
Get order by ID
`shopify_get_order_by_id` * **Purpose:** Fetch details for a specific Shopify order. * **Key Params:** `order_id: str` * **When/How to Use:** Use when a user asks about the status or details of a specific order and provides an order number or ID. Instruct the AI to ask for the order ID if the user's query requires it but isn't provided initially. * **Important Note:** Customers may confuse the order ID with the order name, or some orders may have prefixes. If a lookup fails, ask the customer to double-check that they're providing a valid order ID (not order name) and verify the reference format in their Shopify admin. * **Example Prompt Instruction:** > If the user is asking about the status or location of their order, you must ask them for their order number. Once the user provides an order number, you must call shopify\_get\_order\_by\_id with the provided order\_id as the parameter. Use the information returned by the tool to answer the user's question about their order.Get order history by email
`shopify_get_order_history` * **Purpose:** Retrieve the full order history (most recent 5 orders) for a customer by email address. * **Key Params:** `email: str` * **When/How to Use:** Use when a customer wants to see their recent order history or when you need to look up multiple orders for a customer. This is particularly useful for customers who can't remember specific order numbers. * **Example Prompt Instruction:** > If the user wants to see their recent orders or order history, ask for their email address. Once provided, call shopify\_get\_order\_history with the email parameter to retrieve their most recent 5 orders with full details.Get customer by email
`shopify_get_customer_by_email` * **Purpose:** Look up a Shopify customer by their email address. * **Key Params:** `email: str` * **When/How to Use:** Use when a user asks about their account, order history, or general customer information and provides their email address. Instruct the AI to request the email if needed. * **Example Prompt Instruction:** > If the user is asking about their customer account or looking for information related to orders associated with their account but has not provided an order number, you must ask for the email address associated with their account for verification. Once the user provides their email address, you must call shopify\_get\_customer\_by\_email with the provided email. Use the information returned to help the user with their account-related query.Get product variants
`shopify_get_product_variants` * **Purpose:** Retrieve all variants for specified product IDs, including size, color, and other variant options. * **Key Params:** `product_ids: List[str]` - A list of product IDs to get variants for * **When/How to Use:** Use when customers ask about available options for specific products, or when you need variant information for inventory or pricing queries. * **Example Prompt Instruction:** > When a user asks about available sizes, colors, or other options for a specific product, use shopify\_get\_product\_variants with the product ID(s) to show all available variants and their details.Get all products
`shopify_get_products` * **Purpose:** Retrieve product catalog information including titles, IDs, and tags. Returns up to 3 pages of products sorted by creation date. * **Key Params:** None (retrieves all products) * **When/How to Use:** Use when customers want to browse the product catalog, or when you need to find products by name or category before getting specific details. * **Example Prompt Instruction:** > If a user wants to browse available products or asks "what do you sell?", call shopify\_get\_products to retrieve the product catalog and present the available items.Get fulfillment information
`shopify_get_fulfillment_for_order_id` * **Purpose:** Retrieve fulfillment and shipping information for a specific order ID. * **Key Params:** `order_id: str` - The numeric order ID (e.g., "11607431610736") * **When/How to Use:** Use when customers ask about shipping status, tracking information, or fulfillment details for their order. * **Example Prompt Instruction:** > When a user asks about shipping status or tracking information, first get their order using shopify\_get\_order\_by\_id, then use shopify\_get\_fulfillment\_for\_order\_id with the order ID to get detailed fulfillment information.Get variant metafields
`shopify_get_variant_metafields` * **Purpose:** Fetch custom metafields for product variants, useful for accessing additional product data like specifications, custom attributes, or extended information. * **Key Params:** * `variant_ids: List[str]` - List of variant IDs * `namespace: str` (optional) - Filter by metafield namespace * `key: str` (optional) - Filter by specific metafield key * `type_: str` (optional) - Filter by metafield type * `limit: int` (optional, default 50) - Results per page * `max_pages: int` (optional, default 1) - Maximum pages to fetch * **When/How to Use:** Use when you need access to custom product variant data that's stored in metafields, such as technical specifications, care instructions, or custom attributes. * **Example Prompt Instruction:** > When users ask about detailed product specifications or custom attributes that aren't in the standard product data, use shopify\_get\_variant\_metafields to access additional information stored in metafields.Get product metafields
`shopify_get_product_metafields` * **Purpose:** Fetch custom metafields for products, providing access to additional product information beyond standard fields. * **Key Params:** * `product_ids: List[str]` - List of product IDs * `namespace: str` (optional) - Filter by metafield namespace * `key: str` (optional) - Filter by specific metafield key * `type_: str` (optional) - Filter by metafield type * `limit: int` (optional, default 50) - Results per page * `max_pages: int` (optional, default 1) - Maximum pages to fetch * **When/How to Use:** Use when you need to access custom product data stored in metafields, such as detailed descriptions, specifications, or custom business data. * **Example Prompt Instruction:** > For detailed product information beyond the standard fields, use shopify\_get\_product\_metafields to retrieve custom data stored in metafields for the requested products.Search products
`shopify_search_products` * **Purpose:** Search the product catalog using GraphQL search API to find products matching specific search terms. * **Key Params:** * `search_query: str` - Search terms (e.g., "brake pads", "red shirt", "iPhone case") * `limit: int` (optional, default 10, max 250) - Maximum results to return * **When/How to Use:** Use when customers search for products by name, category, or description. More powerful than browsing all products as it uses Shopify's search capabilities. * **Example Prompt Instruction:** > When users search for specific products or ask about items by name/category, use shopify\_search\_products with their search terms to find matching products with details including pricing, availability, and images.Generate product carousel
`shopify_generate_carousel` * **Purpose:** Generate a product carousel using Shopify's public Storefront API. Can search for products or display specific products by ID. No authentication required. * **Key Params:** * `shop_name: str` - Shopify shop name (e.g., "mystore" for mystore.myshopify.com) * `search_query: str` (optional) - Search term for products * `product_ids: List[str]` (optional) - Specific product IDs to display * `limit: int` (optional, default 6, max 20) - Maximum products for search results * **When/How to Use:** Use to create visual product displays for customers, either based on search results or specific product recommendations. * **Example Prompt Instruction:** > To show customers a visual carousel of products, use shopify\_generate\_carousel with either a search query for discovery or specific product IDs for recommendations. This creates a rich UI display with images and purchase links.Create return/RMA
`shopify_create_return` * **Purpose:** Create a return merchandise authorization (RMA) for a specific order with selected line items. * **Key Params:** * `order_id: str` - Shopify order ID * `line_items: List[ReturnLineItem]` - Items to return with fulfillmentLineItemId and quantity * `reason: str` (optional, default "UNKNOWN") - Return reason (COLOR, DEFECTIVE, NOT\_AS\_DESCRIBED, OTHER, SIZE\_TOO\_LARGE, SIZE\_TOO\_SMALL, STYLE, UNKNOWN, UNWANTED, WRONG\_ITEM) * `notify_customer: bool` (optional, default False) - Whether to email customer * **When/How to Use:** Use when customers want to return specific items from an order. Requires fulfillment line item IDs from the order details. * **Example Prompt Instruction:** > When a customer wants to return items, first get their order details, then use shopify\_create\_return with the specific line items they want to return and the reason for the return.Create replacement order
`shopify_create_return_order` * **Purpose:** Create a 100% discounted replacement order for returned items, typically used in exchange scenarios. * **Key Params:** * `original_order_id: str` - Original order being replaced * `replacement_line_items: List[ReplacementOrderLineItem]` - New items with variantId and quantity * `customer_email: str` - Customer's email address (required) * **When/How to Use:** Use when processing exchanges or sending replacement items at no charge to the customer. * **Example Prompt Instruction:** > For product exchanges or free replacements, use shopify\_create\_return\_order to create a fully discounted order with the replacement items for the customer.Refund order
`shopify_refund_order` * **Purpose:** Process a full refund for a Shopify order with customer email validation for security. * **Key Params:** * `order_id: str` - Order identifier (name like "#1001" or numeric ID) * `customer_email: str` - Customer email for validation * `reason: str` (optional, default "other") - Refund reason ("duplicate", "fraud", "other") * `note: str` (optional) - Note explaining the refund * `notify_customer: bool` (optional, default True) - Send email notification * **When/How to Use:** Use when customers request refunds. Always validates that the order belongs to the provided email address for security. * **Example Prompt Instruction:** > For refund requests, use shopify\_refund\_order with both the order ID and customer email to securely process the full refund. The system will validate ownership before processing.Cancel order
`shopify_cancel_order` * **Purpose:** Cancel a Shopify order with customer email validation, optionally processing automatic refunds. * **Key Params:** * `order_id: str` - Order identifier (name like "#1001" or numeric ID) * `customer_email: str` - Customer email for validation * `reason: str` (optional, default "customer") - Cancellation reason ("customer", "fraud", "inventory", "declined", "other") * `send_notification: bool` (optional, default True) - Send email notification * `process_refund: bool` (optional, default True) - Automatically refund paid orders * `cancellation_note: str` (optional) - Note explaining cancellation * **When/How to Use:** Use when customers want to cancel unfulfilled orders. Cannot cancel already fulfilled orders - use refund instead. * **Example Prompt Instruction:** > For order cancellations, use shopify\_cancel\_order with the order ID and customer email. The system validates ownership and can automatically process refunds for paid orders. Use refund tool for already fulfilled orders.How do I index my entire product site? I'm having trouble with the app finding specific product pages.
We recommend using the “product search” action (see Customize > create new action > Shopify > product search). This will allow the bot to use your Shopify to search for the product information instead, which enables real time information rather than indexing your pages via the web crawler.Can I change the name of my Slack bot, depending on the channel it’s in?
No, the Slack bot’s name is tied to the app - so changing its name will change the name for all bots in your workspace. However, you can still specify which bots are operating in which channels - so you’ll be able to ensure that only your HR bot is responding in the HR channel, etc.Can I train the bot on Slack conversations/channels?
No, although this was a feature, some integral Slack app changes have disabled this feature for now. We're working to re-add this feature soon. Feel free to get in touch to let us know you're interested in it and we will let you know when it's ready.What kind of Slack account do I need in order to connect eesel?
You do not need a corporate Slack account - as long as you have the required permissions to install eesel in their Slack workspace, you can use the Slack integration.
Does the Confluence app or Confluence pages auto-sync?
Confluence app integration: Yes, the Confluence app integration will automatically sync your pages when a page is updated. However, for deleting or archiving pages you will need to manually resync for the changes to appear in eesel. Confluence pages added via extension: No, these pages are indexed as they are when they are initially added. You will need to re-index these pages with the extension when they are updated. If pages are frequently updated, we recommend using our app instead.
Do you integrate with WhatsApp?
If your WhatsApp is connected to your helpdesk, we can likely integrate with it via the helpdesk. To confirm, please let us know whether you have connected WhatsApp to a helpdesk and what that platform is.Do you integrate with N8N?
At the moment, we do not have a default integration with n8n. However, it’s possible you could use our custom API add-on to enable this integration (see our [Custom Actions/APIs](/legacy/integrations/new-or-custom-integration/custom-actions-apis) page for more details), or we could build a custom integration for you. For more details, please send through a description of what kind of integration you’re looking for and the use cases.Do you integrate with social channels like Facebook, Instagram, Meta messenger and other social messaging or commenting platforms?
At the moment, we do not have a default AI Agent integration with social channels. However, it’s possible that if your social messages are coming through to a helpdesk platform you can use our helpdesk integration to reply automatically and handle them there. Please let us know if you want to handle your social messages or comments via a helpdesk, and we can discuss your options. You can use our [Copilot](/legacy/products/ai-copilot) tools to help handle social channel conversations.Preview of custom action tool (subject to change)
| Fields | Description |
|---|---|
| Name | Name of your custom action. This will determine how the action can be called inside of your custom prompt. The name will be automatically converted to custom_api_<name>. |
| URL | Endpoint you want to make a request to. Query parameters are supported. |
| Payload Template (Optional) | Request body to be sent to the endpoint. Can be empty. |
| Headers (Optional) | Request headers to be sent to the endpoint. Can be empty. |
Team Plan - $299/mo
The Team plan is ideal for small teams who are looking for a Copilot solution trained on their website or wiki Docs. You can plug eesel into your Slack, create up to 3 bots, and have 1,000 interactions per month. This plan does not include automatic replies to tickets or triaging actions.Business Plan - $799/mo
The Business Plan is our most popular plan and contains everything from the Team plan but also automatic replies, internal notes, and triaging inside helpdesks. It also includes unlimited bots, and greater integration and training options. You're also provided up to 3,000 interactions per month. This plan is best for teams who want an AI Agent in their helpdesk, or deeper integrations across multiple platforms like Shopify actions.Custom Plan
A Custom Plan is priced around what *you* need. For example, if you need more interactions, or want to set up an integration with a custom platform or one that's not available by default. Custom Plan quotes are prepared based on what you're looking for. For a Custom Plan quote, please prepare a detailed description about what actions, integrations, and usage you're looking for and let the team know at
The export button isn't working properly.
Please use Chrome to export your replies. if this doesn't fix the issue, the please let us know what issue you are experiencing.How to use the ‘add column’ feature in the History tab? What is it for?
The add column feature allows you to create a data column for each reply thread. You can describe to the AI what information it should extract from the conversation in order to populate that column. With this information, you can export your replies and filter/transform the data for specific insights or information. For example, create an "email" column, and ensure that your prompt tells the bot to collect the user's email at the start of the conversation. This way, you will have a list of leads or email addresses from each conversation. Or, you can create a column that is for what the 'topic' is about. You can either ask the AI to generate its own topics by selecting a 'Text' field, or give it topics to choose from.| Privacy Policy and Terms | https://www.eesel.ai/privacy | |
| Trust Center | https://app.aus.vanta.com/eesel.ai/trust/5ncypy4hkv10s9yf7il8sp | |
| Cloud Service Agreement | /files/BauWkLacoPdKuBDCxPRp |
Can I enable MFA?
Yes, this is a custom item that we can arrange for you. Please contact our in-app support once on a paid plan.How does eesel handle prompt injection attacks in Confluence?
The LLM is never given a token or API access to look at the confluence info via tools or any means so prompt injection is not possible. Each bot has a separate copy of the data and it can access nothing else, and this copy is maintained programmatically.How do we ensure only certain Confluence spaces are shared with the bot?
The LLM is never given a token or API access to your Confluence environment. It only works with a copy of the data you explicitly share with the bot, and that copy is maintained programmaticallyIs eesel AI HIPPA compliant?
No, eesel is not HIPPA compliant. However, we work with several healthcare companies. We recommend exploring custom data retention options with us - please reach out with your specific data retention requirements for us to discuss further.Does eesel have indemnity insurance?
Yes! Please reach out for more details.I don’t want to go over my interaction limit. Is it possible to put a hard limit/stop the bots from operation once they hit the interaction limit?
Yes, this is possible. Please let us know if you want us to set that up for you.I hit the interaction limit. Will my service be disrupted?
If you are on a trial account, your bots will stop functioning after a brief grace period. There are no extensions for usage limits for trial accounts. You will need to subscribe to a paid plan in order to continue the functionality of your bots. For paid users, it will not be immediately disrupted and you will be asked to contact support in order to update your plan to reflect the appropriate usage. For the first month of a paid plan, we don't charge overages unless in extenuating circumstances. We use this first month to help determine the best billing path forward. Please reach out to support as soon as possible if you're on a paid plan and hit your limit, in order to continue your service.How much do additional interactions cost?
Generally, interactions are priced at $0.15 per interaction past the default amount provided by the plan. For cost effectiveness, we do recommend upgrading to the Business Plan if you're on a Team Plan to make the most of both the extra interactions and features. For bulk interaction amounts, we can provide discounts or feature bundles but this is case-by-case. Please reach out to the team via the in-app messaging system or atI hit my interaction usage limit. What do I do?
If you've hit your interaction usage limit, please review your average usage for the month across all of your bots and reach out to support to let them know what kind of usage you will need moving forward. We'll be able to provide you with a quote and apply the usage changes to your account from there. If you've hit your interaction usage limit on a trial account, there is no way to extend this usage limit. You should still have some leeway to continue interacting with the bot for a short period of time until it is hard limited. You'll need to subscribe to one of our paid plans to continue using the bots after this hard limit is reached.What exactly counts as an interaction?
Interactions are AI responses from the bot/each time the bot processes a query. Each time the AI responds to a user query, it counts as one interaction, regardless of how many actions (such as tagging a ticket, escalating, or routing) are performed within that response. One ticket can have multiple interactions if the AI responds multiple times. If the AI replies several times in one ticket, each reply is counted as one interaction.How do I terminate my subscription via Atlassian?
1. Head to: \How quickly does the bot respond?
It depends on the platform. In the eesel dashboard chat, responses are typically faster than through platform integrations like Zendesk. Around 8–12 seconds is considered normal on integrated platforms, while 3–10 seconds is normal in the dashboard. If you're experiencing issues with bot response time, please reach out and let us know what platform, which bot, and an example question.Where to find instructions
Open your agent and click the **Instructions** tab in the main content area. Instructions are written in plain text (markdown supported).What to include
Good instructions cover four areas: #### 1. Identity and role Tell the agent who it is and what it does. **Good example:** ``` You are a customer support agent for Acme Corp. You help customers with product questions, troubleshooting, and account issues. ``` **Bad example:** ``` You are an AI. ``` #### 2. Response style Define how the agent should communicate. **Good example:** ``` Response guidelines: - Be friendly and professional - Keep answers concise — aim for 2-3 short paragraphs - Use bullet points for step-by-step instructions - Always greet the customer by name if available - End with "Is there anything else I can help with?" ``` **Bad example:** ``` Be nice. ``` #### 3. Handling specific situations Give clear rules for common scenarios. **Good example:** ``` Specific rules: - For billing questions: Check the customer's account status first, then provide relevant information - For bug reports: Ask for the browser/device, steps to reproduce, and any error messages. Then assign to the Engineering team - For feature requests: Thank the customer, tag as "feature-request", and let them know the product team reviews all requests - For cancellation requests: Empathize, ask for the reason, offer alternatives if appropriate, then assign to the Retention team ``` #### 4. Escalation rules Define exactly when the agent should hand off to a human. **Good example:** ``` Escalation rules: - If you are not confident in your answer, assign to Support Tier 2 - If the customer asks to speak with a human, assign immediately - If the customer mentions legal action, assign to the Legal team - If the issue involves a security concern, assign to Security with a high-priority tag - Never attempt to handle refunds over $100 — assign to a manager ``` **Bad example:** ``` Escalate if needed. ```Template by use case
#### Support agent ``` You are a support agent for [Company Name]. You help customers with questions about [Product]. Response style: - Professional and friendly - Concise answers (2-3 paragraphs max) - Use bullet points for instructions - Include relevant links when available When answering: 1. Search your knowledge base for the answer 2. If found, provide a clear, helpful response 3. If not found, let the customer know and assign to a human agent Escalation rules: - Billing disputes → Billing Team - Bug reports → Engineering - Can't find an answer → Support Tier 2 - Customer requests a human → Assign immediately ``` #### Triage agent ``` You are a ticket triage agent for [Company Name]. Your job is to categorize and route incoming tickets — you do NOT reply to customers. For each ticket: 1. Read the content carefully 2. Tag with one category: billing, bug, how-to, feature-request, account 3. Route based on category: - billing → Billing Team - bug → Engineering - how-to → Support Tier 1 - feature-request → Product - account → Support Tier 1 4. Set priority: urgent / high / normal 5. Leave an internal note explaining your categorization ``` #### Knowledge agent ``` You are an internal knowledge assistant for [Company Name]. Team members ask you questions about company policies, processes, and documentation. Guidelines: - Answer based only on the documentation you have access to - If you don't have the answer, say so clearly - Always link to the source document when possible - Keep answers concise and actionable - Use a casual, helpful tone (this is an internal team tool) ```Tips for better instructions
**Be specific.** Vague instructions lead to inconsistent behavior. Instead of "be helpful," define exactly what helpful looks like. **Include examples.** If you want responses formatted a certain way, show an example in the instructions. **Iterate based on testing.** Test in the chat panel, see what the agent gets wrong, and refine instructions accordingly. **Don't overload.** A focused set of clear rules works better than a long list of edge cases. Start simple and add rules as you discover gaps. **Use the chat assistant.** The AI assistant in the right panel can help you write and refine instructions.| Capability | What it does |
|---|---|
| Draft and send ticket replies | Leave internal notes or comments with a drafted reply, or send direct responses to customers - all automatically or when you ask it to. |
| Tag and categorize tickets | Add tags based on sentiment and context, and update ticket fields automatically. |
| Update ticket status | Create new tickets, mark them as solved, close tickets, and more. |
| Escalate | Handover tickets or conversations to your team |
| Multilingual | 80+ languages - no need to translate your help center |
| Confidence-based handling | Tell your agent what to do when it's not condient: high confidence → direct reply; low confidence → internal note for agent |
| Sentiment-based escalation | Take actions, respond differently, or escalate immediately based on sentiment |
| Loop prevention | Escalated after a certain number of interactions or attempts to help |
| Business hours awareness | Set business hours for certain behaviours |
| Tone and voice customization | Fully customizable tone and voice - all in plain text. |
| Capability | What it does |
|---|---|
| Draft and send ticket replies | Leave internal notes or comments with a drafted reply, or send direct responses to customers - all automatically or when you ask it to. |
| Tag and categorize tickets | Add tags based on sentiment and context, and update ticket fields automatically. |
| Update ticket status | Create new tickets, mark them as solved, close tickets, and more. |
| Escalate | Handover tickets or conversations to your team |
| Multilingual | 80+ languages - no need to translate your help center |
| Confidence-based handling | Tell your agent what to do when it's not condient: high confidence → direct reply; low confidence → internal note for agent |
| Sentiment-based escalation | Take actions, respond differently, or escalate immediately based on sentiment |
| Loop prevention | Escalated after a certain number of interactions or attempts to help |
| Business hours awareness | Set business hours for certain behaviours |
| Tone and voice customization | Fully customizable tone and voice - all in plain text. |