# Bots
Source: https://docs.aide.app/dashboard/bots
Answer repetitive questions with bots
You can create chatbots to embed on your website or storefront. When creating a bot, you can give it a name. Bots use Knowledge and Scenarios by default.
Example of what Aide’s chat embed looks like on a website.
Aide bot configuration.
## Customizing your bot
You customize a bot’s description and instructions, chat settings, and add suggested question buttons.
* **Description and instructions** - the description appears under the bot’s name. Instructions are where you can define the bot’s identity, tone, and general behavior.
* **Chat settings** - there are a few settings available, they are as follows:
* **Team** - you can configure which team the bot belongs to,
* **Bot identifier** - what it identifies itself as e.g. (bot, agent, pup, etc.)
* **Greeting messages** - these are the first messages the bots sends to the user when it’s opened.
* **Contact information** - email address or phone number customer will be directed to if the bot is unable to help.
* **Suggested questions** - are the little pill-like buttons displayed in the bot when the user is interacting with it. These expand to full questions that are then answered by the bot. This is helpful if you want to share frequently asked questions for user to ask the bot.
What the suggested questions editor looks like.
What the suggested question buttons look like in the bot when added to your website.
# Conversations
Source: https://docs.aide.app/dashboard/conversations
The conversations page is where you can see the conversations Aide has seen
It consists of 4 sections:
1. **Filters -** You can filter conversations by Open, Sent, Detected (whether a topic has been detected), Not Detected (conversations where no topic was detected), and All. You can also look up individual conversations by ID.
2. **List of conversations** - a list of conversations.
3. **Conversation** - the thread of messages between the customer and the agent. The main benefit of the conversations page is it shows you conversations they way “Aide sees them”, e.g. what topics were detected (if any), which scenarios were executed, whether a draft was generated or not. More importantly, you can provide feedback on topics and scenarios and draft. To reiterate, you can see:
1. what **Topics** were detected or assign a topic, provide feedback (thumbs up or down when you click a topic).
2. which **Scenarios** were applied and provide feedback (thumbs up or down when you click a scenario).
3. if any **Drafts** were generated and provide feedback (thumbs up or down when you click draft).
4. **Context** - if the customer’s email address, phone number or order number is has an existing order or CRM record, it is retrieved and displayed on the right-side of the screen.
An example of the Aide conversations page. With a specific conversation inquiring about an order. You can see the order information retrieved from Shopify on the left.
An example of how feedback can be provided when “topic” is clicked. You click the thumbs up or down buttons and that trains Aide for future messages.
# Knowledge
Source: https://docs.aide.app/dashboard/knowledge
In addition to Scenarios, Aide uses knowledge to create responses
There are different kinds of knowledge in Aide:
* Zendesk or Front Macros
* Products from eCommerce store
* Manually added articles
* Scraped websites
* Existing help center
* Manually added articles
An example of the different knowledge categories in Aide.
* **Website data** - when you create a new Aide account, if you check the use website data checkbox, your website is scraped for FAQs and similar types of content. If your website has products, they are imported separately by integrating your eCommerce platform.
* **Manual articles** - You can add knowledge articles manually by pasting in a document, you can add a title and a link (optional).
Modal for manually adding knowledge articles into Aide.
* **Importing Help Center** - You can automatically import your Zendesk, Gorgias, Intercom, or HelpScoutknowledge base by clicking the import button and pasting in the link.
Modal for auto-importing help center articles.
* **eCommerce Products -** when you connect your eCommerce platform, Aide automatically imports your products with details including the description, sizes, colors, prices, SKUs and stock status, this information can then be used to recommend products and answer availabilty questions. They are listed in a separate category in articles as “Products from store”
Products imported from Aide store as knowledge articles.
Product details imported include name, description, sizes, colors, prices, SKUs and stock status.
* **Zendesk Macros or Front templates** - text from macros or templates is auto-imported as knowledge when you connect either of these platforms to Aide.
# Reports
Source: https://docs.aide.app/dashboard/reports
View Aide performance and feedback metrics
The reports page on the Aide dashboard is organized into 3 main sections:
## **Conversations section**
The conversations section is made of 4 columns:
1. **Topics** - Overall topic detection rate is reported. i.e. how many conversations got at least one topic. Ideally this number would be as high as possible. If it is low, chances are you do not have enough topics, not enough examples, or inaccurate topic names and descriptions. To dive further, you can click the “**No topics**” row, which would show you a list of conversations with no topics. If you intend to setup automations, we recommend having a comprehensive set of topics. To learn more, read (building a taxonomy).
2. **Scenarios** - An automation rate is reported, which reflects the % of conversations with detected topics that had a full or a partial automation applied. They are divided into two subsections **AI Agent Automations** and **Agent Panel Suggestions:**
1. **AI Agent Automations** - count Scenarios that contain an *AI agent* actions.
1. **Full automations -** are Scenarios where the *AI agent* closes a conversation in one-touch. i.e. they include updating the conversation status to “**closed**” by applying a Zendesk or Front macro.
2. **Partial automations -** are Scenarios with *AI agent* actions but leave the conversation open.
2. **Agent Panel Suggestions** - count Scenarios that contain *Agent Panel* actions. These stats relate to Agent Panel usage by human agents.
1. **Applied suggestions** - are Scenarios with **Suggest response** or **Suggest macro** that were applied by human agents (does not include AI created drafts, which are counted in their own **Drafts** section).
2. **Suggested to agents** - are Scenarios with **Suggest response** or **Suggest macro** that were not applied. So a more apt name for this would be “**Suggested but not applied by agents**”.
3. **Drafts** - An insertion rate is reported, this reflects the % of drafts that were inserted by agents.
1. **Insertions** - count of drafts and whether they were inserted or not.
1. **Inserted drafts** - are drafts that were inserted. This is measured by how many times the “Insert” button for AI drafts on the *Agent Panel* is clicked.
2. **Unused drafts -** are drafts that were not inserted by human agents.
2. **Generations** - count of created drafts, and whether they were manually generated or auto-generated.
1. **Auto-generated drafts** - drafts that were created by a Scenario using a **Create draft** action on the *Agent Panel*.
2. **Manually generated drafts** - drafts created by human agents by clicking the “generate” button on the *Agent Panel*.
4. **Bots** - Handle rate reflects % of conversations that were answered by a bot.
1. **Handled conversations** - total number of conversations that involved a bot interaction.
2. **Unresolved conversations** - conversations that involved a bot interaction but were not resolved.
## **Feedback section**
One of the powers of Aide is the ability for your team members to provide feedback as QA and training for the AI system. Feedback is provided by clicking thumbs up or down on the *Conversations page* or the *Agent Panel*. You can provide feedback on Topics, Scenarios, and Drafts.
The metric we’re measuring with feedback is accuracy, it is based on the things you provided feedback on. For example, if you give one thumbs up and one thumbs down, accuracy would be 50%. It is not calculated over all conversations with topics, just the ones where you provide either thumbs up or down.
The columns of the feedback feedback section are:
1. **Topic detection accuracy**
2. **Automation accuracy**
3. **Drafting accuracy**
4. **Bot** **accuracy**
Within each section denoting how many thumbs up or down were provided. The last section, percent of items reviewed represent the % of conversations where feedback was provided.
## **Topics section**
In the topics section you’ll see a list of your topics sorted by volume. You can also see how many examples were added for training and how many topic detections were corrected. If any corrections are made, the topic detection accuracy is reduced.
# Scenarios
Source: https://docs.aide.app/dashboard/scenarios
Scenarios are how Aide takes actions, such as generating a draft for an agent, or taking an action via AI
Scenarios are composed of two pieces: **conditions** and **actions**.
How scenarios work: when a message is received, Aide looks at whether the interaction matches any of the conditions of your scenarios. If they are all true, then the actions are executed. *There are two types of scenarios:*
* **Always-on Scenario** - always-on Scenarios have no conditions, they always apply. They are useful for adding AI instructions that should always be referenced by the AAI in every conversation like, brand, tone, and style. Following is an example of simple always-on instructions. Notice how it says that the scenario is always applied.
A simple example of always-on Scenario. It includes a note above specifying that the scenario is always applied unless you add conditions.
* **Normal Scenario** with conditions. They are only applied when all the conditions listed are true.
## Conditions
Conditions are based on fields for a given conversation. The fields you have available depend on the integrations you have connected. For example, if you use Zendesk, ticket and user fields are available as conditions. If you use Shopify, order information, and tracking information such as whether an order has been shipped, delivered, or which country it’s being shipped to are available as conditions. If you use Salesforce or a different CRM, account or record specific fields are available as conditions.
Each condition is a row structured as follows:
* **Condition** - Select the condition you want. These can be Aide conditions, or conditions based on your connected integrations (Zendesk, Shopify, Salesforce, etc.)
* **Operator** - Select whether that condition should be (is) or shouldn’t be (is not).
* **Value** - this varies based on the selected condition, but can be true or false, or a set of options to choose from.
**Combining conditions:** you can add multiple conditions using the “and” or “or” buttons. For conditions combined with “and” they all have to be true for the scenario to be applied. So if you have two groups of conditions combined with “or”, either group needs to be true for the scenario to apply.
Conditions combined with and: you can have multiple conditions that should evaluate to true for the scenario to apply. Within each “or” group, you can have multiple “and” statements.
Conditions combined with or: you can have multiple sets of conditions in a scenario any of which should evaluate to true for the scenario to apply. Within each “or” group, you can have multiple “and” statements.
### Aide conditions
By default Aide has several conditions available:
* **Is first message** - only applies the scenario if the message is the first message received in a conversation.
* **Topic** - apply the scenario to messages that have the selected topic.
* **Highest Topic** - if multiple topics are detected, this condition allows you to specify the highest topic as a condition for the scenario to run.
* **Topic confidence -** because Aide assigns a conditions score (from 0 to 100%) to every topic it detects, you can use its confidence as a condition for a scenario. Confidence thresholds run from 80% to 97.5% in increments of 2.5%**.**
* **Inbox** - if you have multiple inboxes connected, you can limit the scenario to running in the inboxes you want.
## Integration specific conditions:
**Zendesk Conditions -** ticket and user fields created in Zendesk are synced to Aide during the integration process, making them available to use as conditions for scenarios.
Custom Zendesk ticket fields.
Zendesk ticket fields available as conditions when building an Aide scenario.
This is especially handy if you have a mature Zendesk setup where ticket and user fields are populated when a ticket is created, so you can then create a scenario with the appropriate conditions to handle those tickets.
### Front Conditions
Similar to Zendesk, Front conversation or contact fields can be used as conditions.
### Gorgias Conditions
Ticket tags and statuses, and other ticket fields can be used as conditions.
## **eCommerce Conditions**
These are the most powerful conditions in Aide and are available as soon as you integrate your eCommerce platform. Shopify and WooCommerce are supported. You can also integrate your custom platform if you contact us.
* **Order exists** (true or false) \*\*\*\*- when a customer reaches out Aide checks if their order exists, as long as their order ID, email, or phone number are provided. This condition allows you create scenarios for when an order exists for a customer or if it doesn’t. Available options for this condition are true or false.
* **Order created** (time in hours ago) - If an order is found, Aide is able to look at when it was placed. This condition is helpful things like canceling orders if they were placed recently enough and haven’t been processed yet.
* **Tracking exists** (true or false) **-** For a given order, Aide checks if tracking information is available or not. This condition allows you create scenarios for when a shipping label has been created for an order or not yet. Available options for this condition are true or false.
* **Tracking begun** (time in weeks ago) \*\*\*\*- If tracking exists, Aide is able to look at when the label was created, typically that is the indicator of “tracking begun”.
* **Tracking last updated** (time in weeks ago) - if tracking exists, \*\*\*\* you can decide things like whether an order should be replaced based on whether a shipment is missing and its tracking status is not delivered, and tracking was last updated a week or more ago.
* **Destination country** (multiple countries) - if a shipment was fulfilled, then destination country can be selected when building Scenarios where your shipping policies differ per country.
* **Shipment status** (label\_created, in\_transit, delivered, exception, not\_trackable) - you can add a Scenario condition based on the shipment status.
* **Tracking company** (UPS, Fedex, DHL, etc.) - you can create scenario conditions based on couriers as well!
### Salesforce conditions:
When you connect Salesforce or any other CRM to Aide, contact fields are available to use as conditions.
Salesforce CRM contact fields available for use as Scenario conditions.
## Actions
Scenario actions are only executed if all the conditions for a given scenario are true. For example, you could have a scenario with a certain topic, received in a certain inbox, with a certain custom field value as conditions. Only then would the actions be executed.
Aide has many different actions categorized into AI agent actions and Agent panel actions:
### **AI agent actions**
Aide’s AI agent works behind the scenes based on your Scenarios to solve tickets automatically. When using Aide as an agent, the following actions are available:
Aide AI agent actions.
1. **AI reply** - \*\*\*\*allows you to automatically send responses composed with generative AI. The responses are composed based on knowledge articles, macros (or message templates), and can be steered by combining them with a prompt \*\*instructions action.\*\*
The AI reply action and escalation options when building a scenario.
**Using AI replies safely**: Aide requires configuring scenarios to ensure that it only responds when all conditions are met. This is done by creating a scenario for a given topic and adding conditions to ensure that it only acts on desired messages. This helps ensure Aide only sends generative responses when it is expected to. Additionally, you should pair generate response with the **prompt instructions action** to define how the response should be written or any factors that it should take into account when writing the response.
**Handling escalation (when AI is unable to create a meaningful response):** by default, Aide will verify the generated response before it is sent to ensure that it addresses the customer’s concerns. if it does not (e.g. they ask a question that is not covered by the knowledge and configuration in Aide), then it will be escalated to a human agent.
**There are two types of escalation:**
* **Escalate with internal comment:** Aide will not send a reply to the customer and will only post an internal comment describing why it did not respond.
* **Escalate with hand-off message:** Aide will send a reply to the customer informing them that a member of the support team will get back to them.
1. **Run automation** - Aide’s AI agent can run actions automatically, but the actions are predefined. For example, if you have an eCommerce integration, you can put an order on hold or update a shipping address.
The Run automation action for an eCommerce account when building a scenario.
1. **Send canned response** - is a simple action type that sends a canned response verbatim. It supports Liquid Markup, enables the automated placement of data in responses.
Example of a canned response that uses Liquid Markup and available data. See appendix for an example of available data object.
1. **Apply macro** - Automatically applies a macro (set of actions) in Zendesk or Front. Macros are help desk shortcuts that when executed can resnd a response, or atake any other type of action, like tagging, assigning conversations, setting priority, and so on. If you want to perform any type of help desk action, chances are it is done using a macro. When you connect Zendesk or Front, Aide automatically imports your macros and they can be applied by scenarios.
* **Post internal comment** - Use this to post a piece of text as an internal comment.
* **Collect field -** If there is certain information you need to collect from a given message, like a date or a number, you can have Aide automatically collect that and save it to a field of your choice. You may need our team’s assistance configuring this.
* **AI instructions** - Is one of the most powerful actions, especially handy when you’re using AI replies or creating drafts. This is where you can add instructions for the LLM to follow when composing the response.
### **Agent panel actions**
You can configure actions in the Scenario that are displayed to your human agents in the “Agent Panel” (see Appendix I).
Aide Agent Panel (for human agents) actions.
1. **Create draft** - are composed using generative AI just like the AI reply action, with the only exception being that the responses are not sent to the customer, but rather, are left as drafts for agents to review, insert, and send manually. They are presented in the Aide panel, available in Zendesk and Front. If using Gorgias, they will be pasted as internal comments visible only to agents.
2. **Suggest response** - is an action that presents a text response verbatim for agents to see in the Aide panel, available in Zendesk and Front. These suggested responses are independent of generated drafts and are displayed as-is. Like canned responses, they can also use variables. Suggest response can be paired with **generate response** or **pre-generate draft** actions to be used as a base for composing a response with generative AI. You can also add a **prompt instructions** action to modify the generated response further if you choose to.
3. **Suggest macro** - presents a macro for agents to see in the Aide panel, available in Zendesk and Front. These suggested macros are displayed independently and can be applied by an agent in 1-click.
# Simulator
Source: https://docs.aide.app/dashboard/simulator
The simulator is the easiest way to test Aide
It is a chat where you can ask Aide questions and see how it would respond. The goal is for you to role-play as a customer as you edit your Aide setup. It is divided into 3 sections:
1. **Conversations list** - list of conversations.
2. **Conversation thread** - this is identical to the conversations page, where once you send a message, Aide detects topic, applies any relevant scenarios, searches across knowledge, and responds accordingly.
3. **Context** - if you use ticket fields or user fields on Zendesk you can enter mock values before starting a simulated conversation to test conditional Scenarios. The simulator can also answer questions about your products if your eCommerce platform is connected to Aide.
# Topics
Source: https://docs.aide.app/dashboard/topics
Auto-sort every incoming message into a 'bucket'
Aide trains and finetunes a classifier for every customer account based on their data. This is essential for the categorization of incoming conversations, so that when it comes in it gets sorted in a bucket for further actions.
Aide creates topics automatically when you connect your ticket source (such as Front, Zendesk, Gorgias, or Gmail to Aide). We encourage teams to edit topics, create new ones to ensure maximal coverage of incoming tickets.
Coverage means the percentage of conversations that are identified as belonging to the existing topics. This represents the percentage of conversation where at least one topic is detected.
Topics are organized in a 3 level hierarchy called a “taxonomy” it is a list of topics organized by category and group. You can create new categories, new groups within those categories, and new topics under each group.
Create a topic or a new category using the “Create” button on the [Topics page](https://admin.aide.app/topics).
Edit categories or create new subgroups under them by hovering over the 3 dots menu.
Edit groups or create new topics under them by hovering over the 3 dots menu.
The benefit of detecting topics in incoming messages is that you can define how each topic is handled (via **Scenarios**), whether through automation or suggestions for agents.
## How to train topics and improve detection
Aide uses a topic’s name, description, and examples to determine whether a conversation belongs to that topic or not. For the following examples, let’s assume we have a topic named “**Book review**”, with a description of “**Customer provides the review of a book they recently read”**, and have the following examples:
You can add examples through the Aide dashboard by going into the topic and manually adding the example. Or through the agent copilot in Zendesk or Front (*see Agent Panel documentation for instructions on how this is done*), by providing feedback (thumbs up or down) adds a positive or negative example respectively, the two types of examples:
* **Positive examples** - when training, if you give a message thumbs up, Aide will add the message as a positive example. When teaches Aide to choose that topic for similar messages in the future. Positive examples are unmarked.
What a positive example looks like.
* **Negative examples** - when training if you give a message thumbs down, Aide will add the message as a “negative example”, which teaches Aide not to choose that topic for similar messages in the future. Negative examples are highlighted in red and are noted as such.
What a negative example looks like.
**Automated review of examples**
To make your life easier, when a new topic is created. This happens regardless of whether a topic is created by Aide during initial setup or by you after that, this helps build up training examples for the topic to improve accuracy.
* **Suggested examples** - Aide automatically scans new conversations for examples for existing topics and adds them as suggestions for you to approve or rejected.
* **Outlier examples** - Aide automatically reviews examples present in every topic, even if they were manually added by your team, to ensure that they match the topic name, description, and other examples. In the following example, the message is asking for an update on their order, but as noted above, the topic we are looking at is “Book views”, so Aide is suggesting that it be moved to the correct topic, “**Order tracking**” and is presenting a button for you to do click and it will do so. You also choose keep it where it is by clicking “keep”.
An example that does not belong to “Book reviews”, Aide wants to move it to the “Order tracking” topic.
More importantly, when new examples are added, Aide compares them with the existing examples to determine if they really belong in that topic. If they don’t, Aide will suggest moving it to a more fitting topic. This is to ensure the consistency training examples and maintain the topic integrity for precise and accurate detection.
## **Managing existing examples**
For a given example, you can either switch it from positive to negative (or vice versa), move it to a different topic, or delete it entirely.
# Custom integration
Source: https://docs.aide.app/developers/custom-integration
Connecting a custom integration to Aide
### Custom integrations
This document provides a simple OpenAPI specification for the minimum API requirements needed to connect your data systems with Aide. We maintain flexibility regarding the fields returned in responses, requiring only a server endpoint with a URL and API token that implements three core endpoints: listing latest orders, retrieving individual order details, and fetching customer information.
````openapi: 3.0.0
info:
title: Order & Customer Management Integration API
version: '1.0.0'
description: |
Specification for integrating order and customer management platforms with Aide AI.
This API defines the minimum required endpoints and data structures
while allowing flexibility in implementation details.
Aide AI will poll these endpoints regularly (approximately once per minute)
to sync order and customer data.
servers:
- url: https://api.{platform}.com/v1
variables:
platform:
default: yourplatform
description: Your platform's API domain
paths:
/latest_orders:
get:
summary: List most recently updated orders
description: |
Retrieve the most recently updated orders, sorted by last update time in descending order.
This endpoint will be polled approximately once per minute by Aide AI to detect changes.
Implementation must provide at least the order ID and last update time.
parameters: - in: query
name: page_size
schema:
type: integer
minimum: 1
maximum: 100
default: 50
required: false
description: Number of orders to return per page - in: query
name: page_token
schema:
type: string
required: false
description: Token for pagination. Can be implemented as cursor or page number
responses:
'200':
description: Successfully retrieved orders
content:
application/json:
schema:
type: object
required: - orders
properties:
orders:
type: array
items:
$ref: '#/components/schemas/OrderSummary'
next_page_token:
type: string
description: Token for retrieving the next page of results
/orders/{orderId}:
get:
summary: Get order details
description: Retrieve detailed information about a specific order
parameters: - name: orderId
in: path
required: true
schema:
type: string
responses:
'200':
description: Successfully retrieved order details
content:
application/json:
schema:
$ref: '#/components/schemas/OrderDetail'
'404':
description: Order not found
/customers/{customerId}:
get:
summary: Get customer details
description: Retrieve detailed information about a specific customer
parameters: - name: customerId
in: path
required: true
schema:
type: string
responses:
'200':
description: Successfully retrieved customer details
content:
application/json:
schema:
$ref: '#/components/schemas/CustomerDetail'
'404':
description: Customer not found
components:
schemas:
OrderSummary:
type: object
required: - id - updated_at
properties:
id:
type: string
description: Unique identifier for the order
updated_at:
type: string
format: date-time
description: Time the order was last updated # Additional optional fields can be included
OrderDetail:
type: object
required:
- id
- created_at
- total_amount
properties:
id:
type: string
description: Unique identifier for the order
display_id:
type: string
description: Human-readable order identifier (e.g., order number, confirmation number)
created_at:
type: string
format: date-time
description: Time the order was placed
updated_at:
type: string
format: date-time
description: Time the order was last updated
total_amount:
type: number
format: float
description: Total amount for the order
currency:
type: string
description: Currency code (e.g., USD, EUR)
canceled_at:
type: string
format: date-time
nullable: true
description: Time the order was canceled, if applicable
customer_id:
type: string
description: ID of the customer who placed the order
status:
type: string
description: Current status of the order
billing_address:
type: object
description: |
Billing address as a JSON object. Common fields include:
- address1: Primary street address
- address2: Secondary street address (apt, suite, etc)
- city: City name
- province/state: Province or state
- zip/postal_code: ZIP or postal code
- country: Country code (e.g., US, GB)
You may include additional fields as needed for your platform
# Additional fields can be included
CustomerDetail:
type: object
required:
- id
properties:
id:
type: string
description: Unique identifier for the customer
first_name:
type: string
last_name:
type: string
email:
type: string
format: email
phone:
type: string
country:
type: string
description: Country code (e.g., US, GB)
billing_address:
type: object
description: |
Billing address as a JSON object. Common fields include:
- address1: Primary street address
- address2: Secondary street address (apt, suite, etc)
- city: City name
- province/state: Province or state
- zip/postal_code: ZIP or postal code
- country: Country code (e.g., US, GB)
You may include additional fields as needed for your platform
# Additional fields can be included
securitySchemes:
ApiKeyAut```
---
````
# Introduction
Source: https://docs.aide.app/index
Welcome to the Aide documentation
These docs detail the features offered in the dashboard and how to use them. For more advanced tips and recommendations, reach out to the Aide team for guidance specific to your use-case or industry.
**Aide features that work out the box:**
* **One click integrations**- for popular support and eCommerce and CRM platforms — Zendesk, Gorgias, Front, Shopify, WooCommerce, and Salesforce, with the ability to integrate with custom APIs as needed. See the next page for information on integrations.
* **Customer identification** - automatic retrieval of customer information from your CRM or eCommerce platform based on their email, phone number, order ID, or any other unique identifier.
* **Knowledge based drafts** - based on previous conversations, existing macros and canned templates, and your help center, enriched with real customer data from your CRM or eCommerce platform.
* **Topic suggestions** - Aide suggests a set of topics to help you get started. Topics categorize incoming messages for reporting and creating action-based automated scenarios as opposed to relying solely on your help center and response templates.
For use cases where customers want to simply get AI drafts, they can rely on knowledge articles.
For cases where accuracy is paramount and automation pathways and workflows can be defined, a structured approach using topics and scenarios is recommended.
# Integrations
Source: https://docs.aide.app/integrations/index
Connect Aide to your systems
Aide has several integrations. They are all one-click and connect in under 5 minutes. Zero engineering required.
### One-click integrations
* **Front** - for information about Aide’s Front integration, go [here](https://front.com/integrations/aide).
* **Zendesk** - for information about Aide’s Zendesk integration, go [here](https://www.zendesk.com/marketplace/apps/support/421826/aide-ai-support-automation/).
* **Gorgias** - for information about Aide’s Gorgias integration, go [here](https://www.gorgias.com/apps/aide-support).
* **Shopify** - for information on Aide’s Shopify integration, go [here](https://apps.shopify.com/aide).
* **Wooocommerce** - Aide’s WooCommerce is identical to Shopify integration.
* **Salesforce** - Aide’s Salesforce integration is documented separately.
### Custom integrations
# [See docs for integrating Aide here](/developers/custom-integration)
***
# Agent Panel
Source: https://docs.aide.app/products/agent-panel
Accelerate your agents with real-time AI
The Agent Panel is a tool helpful for teams that use Aide with human agents. It is available on Zendesk and Front. It presents real-time information for every conversation or ticket you view. It is composed of 4 sections:
1. **Context** - if the customer’s email address, phone number or order number is has an existing order or CRM record, it is retrieved and displayed as the first thing.
2. **Topic** - shows what topic was detected (if any). You can provide feedback with the thumbs up or down buttons. If no topics were detected, you can select a topic from the dropdown menu then click the checkmark button that appears.
3. **Generated draft** - if you’re using create draft in a Scenario that applied, you will find a pre-generated draft. If any knowledge articles were used they would be cited. To use the draft, click the Insert button, which will insert it into your Zendesk or Front composer. You can also provide feedback on the draft using the thumbs up or down buttons.
4. **Macros or suggested responses** - if there any Scenarios with suggested macros or responses, they would appear here.
As a last resort, potentially relevant macros or actions are listed towards the bottom.
Aide Agent Panel open in Zendesk.
Aide Agent Panel open in Front.
# AI Agent
Source: https://docs.aide.app/products/ai-agent
Automatically respond to customers and perform actions using AI
Aide's AI agent works behind the scenes to automatically handle customer conversations and perform actions based on your configured Scenarios. The AI agent can respond to customers using generative AI and automate various help desk and business operations.
Aide AI agent actions available in Scenarios.
## How the AI agent works
The AI agent operates through **Scenarios** - combinations of conditions and actions that determine when and how the AI should respond. When a message is received, Aide evaluates whether it matches your scenario conditions, and if so, executes the configured actions.
The AI agent relies on several sources of knowledge to create accurate responses:
* **Knowledge articles** - imported help center content, manually added articles, and scraped website data
* **Macros and templates** - automatically imported from your help desk platform
* **Product information** - automatically synced from your eCommerce platform
* **AI instructions** - custom prompts and guidelines defined in your Scenarios
## AI agent actions
When using Aide as an AI agent, the following actions are available in Scenarios:
### AI Reply
The **AI reply** action allows you to automatically send responses composed with generative AI. Responses are created based on your knowledge base, macros, and can be guided by AI instructions.
The AI reply action and escalation options when building a scenario.
**Safety features:** Aide requires configuring scenarios with specific conditions to ensure the AI only responds when appropriate. This prevents the AI from sending responses to unexpected messages.
**Response verification:** By default, Aide verifies that generated responses fully address customer questions before sending them. If verification fails, the conversation is escalated to a human agent.
The options for escalation are as follows:
* **Escalate with internal comment:** Aide will not send a reply to the customer and will only post an internal comment describing why it did not respond.
* **Escalate with hand-off message:** Aide will send a reply to the customer informing them that a member of the support team will get back to them.
## Escalation handling
When the AI agent cannot create a meaningful response or encounters issues, it escalates to human agents using one of two methods:
* **Escalate with internal comment** - No customer response is sent; only an internal comment explains why the AI didn't respond
* **Escalate with hand-off message** - A customer-facing message informs them that a support team member will follow up
### Run Automation
Aide's AI agent can automatically execute predefined actions, such as:
* **eCommerce actions** - putting orders on hold, updating shipping addresses, canceling orders
* **Help desk actions** - applying tags, changing ticket status, assigning conversations
* **CRM actions** - updating contact information, creating records
The Run automation action for an eCommerce account when building a scenario.
### Send Canned Response
Send predefined responses verbatim, with support for Liquid Markup to dynamically insert customer and order data.
### Apply Macro
Automatically execute help desk macros (Zendesk/Front) or templates to perform actions like tagging, assigning, or changing ticket properties.
### Other Actions
* **Post internal comment** - Add internal notes for your team
* **Collect field** - Automatically extract and save specific information from messages
* **AI instructions** - Provide custom guidance for the AI when composing responses
## Enhanced safety and control
AI agent actions are only executed when all scenario conditions are met, ensuring predictable behavior. For regulated industries or situations requiring high accuracy:
1. **Use Topics and Scenarios** instead of relying solely on past responses and knowledge
2. **Define specific conditions** that must be true before actions are taken
3. **Test scenarios** using the [Simulator](https://my.aide.app/simulator) to verify AI behavior
This deterministic approach helps prevent AI hallucinations and ensures the agent only acts when explicitly configured to do so.
## Getting started
To set up your AI agent:
1. **Configure Topics** to categorize incoming messages
2. **Create Scenarios** with conditions and AI agent actions
3. **Import Knowledge** from your help center and add custom articles
4. **Test scenarios** using the simulator before going live
Example of what Aide’s chat embed looks like on a website.
Aide bot configuration.
## Customizing your bot
You customize a bot’s description and instructions, chat settings, and add suggested question buttons.
* **Description and instructions** - the description appears under the bot’s name. Instructions are where you can define the bot’s identity, tone, and general behavior.
* **Chat settings** - there are a few settings available, they are as follows:
* **Team** - you can configure which team the bot belongs to,
* **Bot identifier** - what it identifies itself as e.g. (bot, agent, pup, etc.)
* **Greeting messages** - these are the first messages the bots sends to the user when it’s opened.
* **Contact information** - email address or phone number customer will be directed to if the bot is unable to help.
* **Suggested questions** - are the little pill-like buttons displayed in the bot when the user is interacting with it. These expand to full questions that are then answered by the bot. This is helpful if you want to share frequently asked questions for user to ask the bot.
What the suggested questions editor looks like.
An example of the Aide conversations page. With a specific conversation inquiring about an order. You can see the order information retrieved from Shopify on the left.
An example of how feedback can be provided when “topic” is clicked. You click the thumbs up or down buttons and that trains Aide for future messages.
# Knowledge
Source: https://docs.aide.app/dashboard/knowledge
In addition to Scenarios, Aide uses knowledge to create responses
There are different kinds of knowledge in Aide:
* Zendesk or Front Macros
* Products from eCommerce store
* Manually added articles
* Scraped websites
* Existing help center
* Manually added articles
An example of the different knowledge categories in Aide.
* **Website data** - when you create a new Aide account, if you check the use website data checkbox, your website is scraped for FAQs and similar types of content. If your website has products, they are imported separately by integrating your eCommerce platform.
* **Manual articles** - You can add knowledge articles manually by pasting in a document, you can add a title and a link (optional).
Modal for manually adding knowledge articles into Aide.
* **Importing Help Center** - You can automatically import your Zendesk, Gorgias, Intercom, or HelpScoutknowledge base by clicking the import button and pasting in the link.
Modal for auto-importing help center articles.
* **eCommerce Products -** when you connect your eCommerce platform, Aide automatically imports your products with details including the description, sizes, colors, prices, SKUs and stock status, this information can then be used to recommend products and answer availabilty questions. They are listed in a separate category in articles as “Products from store”
Products imported from Aide store as knowledge articles.
Product details imported include name, description, sizes, colors, prices, SKUs and stock status.
* **Zendesk Macros or Front templates** - text from macros or templates is auto-imported as knowledge when you connect either of these platforms to Aide.
# Reports
Source: https://docs.aide.app/dashboard/reports
View Aide performance and feedback metrics
The reports page on the Aide dashboard is organized into 3 main sections:
## **Conversations section**
The conversations section is made of 4 columns:
1. **Topics** - Overall topic detection rate is reported. i.e. how many conversations got at least one topic. Ideally this number would be as high as possible. If it is low, chances are you do not have enough topics, not enough examples, or inaccurate topic names and descriptions. To dive further, you can click the “**No topics**” row, which would show you a list of conversations with no topics. If you intend to setup automations, we recommend having a comprehensive set of topics. To learn more, read (building a taxonomy).
2. **Scenarios** - An automation rate is reported, which reflects the % of conversations with detected topics that had a full or a partial automation applied. They are divided into two subsections **AI Agent Automations** and **Agent Panel Suggestions:**
1. **AI Agent Automations** - count Scenarios that contain an *AI agent* actions.
1. **Full automations -** are Scenarios where the *AI agent* closes a conversation in one-touch. i.e. they include updating the conversation status to “**closed**” by applying a Zendesk or Front macro.
2. **Partial automations -** are Scenarios with *AI agent* actions but leave the conversation open.
2. **Agent Panel Suggestions** - count Scenarios that contain *Agent Panel* actions. These stats relate to Agent Panel usage by human agents.
1. **Applied suggestions** - are Scenarios with **Suggest response** or **Suggest macro** that were applied by human agents (does not include AI created drafts, which are counted in their own **Drafts** section).
2. **Suggested to agents** - are Scenarios with **Suggest response** or **Suggest macro** that were not applied. So a more apt name for this would be “**Suggested but not applied by agents**”.
3. **Drafts** - An insertion rate is reported, this reflects the % of drafts that were inserted by agents.
1. **Insertions** - count of drafts and whether they were inserted or not.
1. **Inserted drafts** - are drafts that were inserted. This is measured by how many times the “Insert” button for AI drafts on the *Agent Panel* is clicked.
2. **Unused drafts -** are drafts that were not inserted by human agents.
2. **Generations** - count of created drafts, and whether they were manually generated or auto-generated.
1. **Auto-generated drafts** - drafts that were created by a Scenario using a **Create draft** action on the *Agent Panel*.
2. **Manually generated drafts** - drafts created by human agents by clicking the “generate” button on the *Agent Panel*.
4. **Bots** - Handle rate reflects % of conversations that were answered by a bot.
1. **Handled conversations** - total number of conversations that involved a bot interaction.
2. **Unresolved conversations** - conversations that involved a bot interaction but were not resolved.
## **Feedback section**
One of the powers of Aide is the ability for your team members to provide feedback as QA and training for the AI system. Feedback is provided by clicking thumbs up or down on the *Conversations page* or the *Agent Panel*. You can provide feedback on Topics, Scenarios, and Drafts.
The metric we’re measuring with feedback is accuracy, it is based on the things you provided feedback on. For example, if you give one thumbs up and one thumbs down, accuracy would be 50%. It is not calculated over all conversations with topics, just the ones where you provide either thumbs up or down.
The columns of the feedback feedback section are:
1. **Topic detection accuracy**
2. **Automation accuracy**
3. **Drafting accuracy**
4. **Bot** **accuracy**
Within each section denoting how many thumbs up or down were provided. The last section, percent of items reviewed represent the % of conversations where feedback was provided.
## **Topics section**
In the topics section you’ll see a list of your topics sorted by volume. You can also see how many examples were added for training and how many topic detections were corrected. If any corrections are made, the topic detection accuracy is reduced.
# Scenarios
Source: https://docs.aide.app/dashboard/scenarios
Scenarios are how Aide takes actions, such as generating a draft for an agent, or taking an action via AI
Scenarios are composed of two pieces: **conditions** and **actions**.
How scenarios work: when a message is received, Aide looks at whether the interaction matches any of the conditions of your scenarios. If they are all true, then the actions are executed. *There are two types of scenarios:*
* **Always-on Scenario** - always-on Scenarios have no conditions, they always apply. They are useful for adding AI instructions that should always be referenced by the AAI in every conversation like, brand, tone, and style. Following is an example of simple always-on instructions. Notice how it says that the scenario is always applied.
A simple example of always-on Scenario. It includes a note above specifying that the scenario is always applied unless you add conditions.
* **Normal Scenario** with conditions. They are only applied when all the conditions listed are true.
## Conditions
Conditions are based on fields for a given conversation. The fields you have available depend on the integrations you have connected. For example, if you use Zendesk, ticket and user fields are available as conditions. If you use Shopify, order information, and tracking information such as whether an order has been shipped, delivered, or which country it’s being shipped to are available as conditions. If you use Salesforce or a different CRM, account or record specific fields are available as conditions.
Each condition is a row structured as follows:
* **Condition** - Select the condition you want. These can be Aide conditions, or conditions based on your connected integrations (Zendesk, Shopify, Salesforce, etc.)
* **Operator** - Select whether that condition should be (is) or shouldn’t be (is not).
* **Value** - this varies based on the selected condition, but can be true or false, or a set of options to choose from.
**Combining conditions:** you can add multiple conditions using the “and” or “or” buttons. For conditions combined with “and” they all have to be true for the scenario to be applied. So if you have two groups of conditions combined with “or”, either group needs to be true for the scenario to apply.
Conditions combined with and: you can have multiple conditions that should evaluate to true for the scenario to apply. Within each “or” group, you can have multiple “and” statements.
Conditions combined with or: you can have multiple sets of conditions in a scenario any of which should evaluate to true for the scenario to apply. Within each “or” group, you can have multiple “and” statements.
### Aide conditions
By default Aide has several conditions available:
* **Is first message** - only applies the scenario if the message is the first message received in a conversation.
* **Topic** - apply the scenario to messages that have the selected topic.
* **Highest Topic** - if multiple topics are detected, this condition allows you to specify the highest topic as a condition for the scenario to run.
* **Topic confidence -** because Aide assigns a conditions score (from 0 to 100%) to every topic it detects, you can use its confidence as a condition for a scenario. Confidence thresholds run from 80% to 97.5% in increments of 2.5%**.**
* **Inbox** - if you have multiple inboxes connected, you can limit the scenario to running in the inboxes you want.
## Integration specific conditions:
**Zendesk Conditions -** ticket and user fields created in Zendesk are synced to Aide during the integration process, making them available to use as conditions for scenarios.
Custom Zendesk ticket fields.
Zendesk ticket fields available as conditions when building an Aide scenario.
This is especially handy if you have a mature Zendesk setup where ticket and user fields are populated when a ticket is created, so you can then create a scenario with the appropriate conditions to handle those tickets.
### Front Conditions
Similar to Zendesk, Front conversation or contact fields can be used as conditions.
### Gorgias Conditions
Ticket tags and statuses, and other ticket fields can be used as conditions.
## **eCommerce Conditions**
These are the most powerful conditions in Aide and are available as soon as you integrate your eCommerce platform. Shopify and WooCommerce are supported. You can also integrate your custom platform if you contact us.
* **Order exists** (true or false) \*\*\*\*- when a customer reaches out Aide checks if their order exists, as long as their order ID, email, or phone number are provided. This condition allows you create scenarios for when an order exists for a customer or if it doesn’t. Available options for this condition are true or false.
* **Order created** (time in hours ago) - If an order is found, Aide is able to look at when it was placed. This condition is helpful things like canceling orders if they were placed recently enough and haven’t been processed yet.
* **Tracking exists** (true or false) **-** For a given order, Aide checks if tracking information is available or not. This condition allows you create scenarios for when a shipping label has been created for an order or not yet. Available options for this condition are true or false.
* **Tracking begun** (time in weeks ago) \*\*\*\*- If tracking exists, Aide is able to look at when the label was created, typically that is the indicator of “tracking begun”.
* **Tracking last updated** (time in weeks ago) - if tracking exists, \*\*\*\* you can decide things like whether an order should be replaced based on whether a shipment is missing and its tracking status is not delivered, and tracking was last updated a week or more ago.
* **Destination country** (multiple countries) - if a shipment was fulfilled, then destination country can be selected when building Scenarios where your shipping policies differ per country.
* **Shipment status** (label\_created, in\_transit, delivered, exception, not\_trackable) - you can add a Scenario condition based on the shipment status.
* **Tracking company** (UPS, Fedex, DHL, etc.) - you can create scenario conditions based on couriers as well!
### Salesforce conditions:
When you connect Salesforce or any other CRM to Aide, contact fields are available to use as conditions.
Salesforce CRM contact fields available for use as Scenario conditions.
## Actions
Scenario actions are only executed if all the conditions for a given scenario are true. For example, you could have a scenario with a certain topic, received in a certain inbox, with a certain custom field value as conditions. Only then would the actions be executed.
Aide has many different actions categorized into AI agent actions and Agent panel actions:
### **AI agent actions**
Aide’s AI agent works behind the scenes based on your Scenarios to solve tickets automatically. When using Aide as an agent, the following actions are available:
Aide AI agent actions.
1. **AI reply** - \*\*\*\*allows you to automatically send responses composed with generative AI. The responses are composed based on knowledge articles, macros (or message templates), and can be steered by combining them with a prompt \*\*instructions action.\*\*
The AI reply action and escalation options when building a scenario.
**Using AI replies safely**: Aide requires configuring scenarios to ensure that it only responds when all conditions are met. This is done by creating a scenario for a given topic and adding conditions to ensure that it only acts on desired messages. This helps ensure Aide only sends generative responses when it is expected to. Additionally, you should pair generate response with the **prompt instructions action** to define how the response should be written or any factors that it should take into account when writing the response.
**Handling escalation (when AI is unable to create a meaningful response):** by default, Aide will verify the generated response before it is sent to ensure that it addresses the customer’s concerns. if it does not (e.g. they ask a question that is not covered by the knowledge and configuration in Aide), then it will be escalated to a human agent.
**There are two types of escalation:**
* **Escalate with internal comment:** Aide will not send a reply to the customer and will only post an internal comment describing why it did not respond.
* **Escalate with hand-off message:** Aide will send a reply to the customer informing them that a member of the support team will get back to them.
1. **Run automation** - Aide’s AI agent can run actions automatically, but the actions are predefined. For example, if you have an eCommerce integration, you can put an order on hold or update a shipping address.
The Run automation action for an eCommerce account when building a scenario.
1. **Send canned response** - is a simple action type that sends a canned response verbatim. It supports Liquid Markup, enables the automated placement of data in responses.
Example of a canned response that uses Liquid Markup and available data. See appendix for an example of available data object.
1. **Apply macro** - Automatically applies a macro (set of actions) in Zendesk or Front. Macros are help desk shortcuts that when executed can resnd a response, or atake any other type of action, like tagging, assigning conversations, setting priority, and so on. If you want to perform any type of help desk action, chances are it is done using a macro. When you connect Zendesk or Front, Aide automatically imports your macros and they can be applied by scenarios.
* **Post internal comment** - Use this to post a piece of text as an internal comment.
* **Collect field -** If there is certain information you need to collect from a given message, like a date or a number, you can have Aide automatically collect that and save it to a field of your choice. You may need our team’s assistance configuring this.
* **AI instructions** - Is one of the most powerful actions, especially handy when you’re using AI replies or creating drafts. This is where you can add instructions for the LLM to follow when composing the response.
### **Agent panel actions**
You can configure actions in the Scenario that are displayed to your human agents in the “Agent Panel” (see Appendix I).
Aide Agent Panel (for human agents) actions.
1. **Create draft** - are composed using generative AI just like the AI reply action, with the only exception being that the responses are not sent to the customer, but rather, are left as drafts for agents to review, insert, and send manually. They are presented in the Aide panel, available in Zendesk and Front. If using Gorgias, they will be pasted as internal comments visible only to agents.
2. **Suggest response** - is an action that presents a text response verbatim for agents to see in the Aide panel, available in Zendesk and Front. These suggested responses are independent of generated drafts and are displayed as-is. Like canned responses, they can also use variables. Suggest response can be paired with **generate response** or **pre-generate draft** actions to be used as a base for composing a response with generative AI. You can also add a **prompt instructions** action to modify the generated response further if you choose to.
3. **Suggest macro** - presents a macro for agents to see in the Aide panel, available in Zendesk and Front. These suggested macros are displayed independently and can be applied by an agent in 1-click.
# Simulator
Source: https://docs.aide.app/dashboard/simulator
The simulator is the easiest way to test Aide
It is a chat where you can ask Aide questions and see how it would respond. The goal is for you to role-play as a customer as you edit your Aide setup. It is divided into 3 sections:
1. **Conversations list** - list of conversations.
2. **Conversation thread** - this is identical to the conversations page, where once you send a message, Aide detects topic, applies any relevant scenarios, searches across knowledge, and responds accordingly.
3. **Context** - if you use ticket fields or user fields on Zendesk you can enter mock values before starting a simulated conversation to test conditional Scenarios. The simulator can also answer questions about your products if your eCommerce platform is connected to Aide.
# Topics
Source: https://docs.aide.app/dashboard/topics
Auto-sort every incoming message into a 'bucket'
Aide trains and finetunes a classifier for every customer account based on their data. This is essential for the categorization of incoming conversations, so that when it comes in it gets sorted in a bucket for further actions.
Aide creates topics automatically when you connect your ticket source (such as Front, Zendesk, Gorgias, or Gmail to Aide). We encourage teams to edit topics, create new ones to ensure maximal coverage of incoming tickets.
Coverage means the percentage of conversations that are identified as belonging to the existing topics. This represents the percentage of conversation where at least one topic is detected.
Topics are organized in a 3 level hierarchy called a “taxonomy” it is a list of topics organized by category and group. You can create new categories, new groups within those categories, and new topics under each group.
Create a topic or a new category using the “Create” button on the [Topics page](https://admin.aide.app/topics).
Edit categories or create new subgroups under them by hovering over the 3 dots menu.
Edit groups or create new topics under them by hovering over the 3 dots menu.
The benefit of detecting topics in incoming messages is that you can define how each topic is handled (via **Scenarios**), whether through automation or suggestions for agents.
## How to train topics and improve detection
Aide uses a topic’s name, description, and examples to determine whether a conversation belongs to that topic or not. For the following examples, let’s assume we have a topic named “**Book review**”, with a description of “**Customer provides the review of a book they recently read”**, and have the following examples:
You can add examples through the Aide dashboard by going into the topic and manually adding the example. Or through the agent copilot in Zendesk or Front (*see Agent Panel documentation for instructions on how this is done*), by providing feedback (thumbs up or down) adds a positive or negative example respectively, the two types of examples:
* **Positive examples** - when training, if you give a message thumbs up, Aide will add the message as a positive example. When teaches Aide to choose that topic for similar messages in the future. Positive examples are unmarked.
What a positive example looks like.
* **Negative examples** - when training if you give a message thumbs down, Aide will add the message as a “negative example”, which teaches Aide not to choose that topic for similar messages in the future. Negative examples are highlighted in red and are noted as such.
What a negative example looks like.
**Automated review of examples**
To make your life easier, when a new topic is created. This happens regardless of whether a topic is created by Aide during initial setup or by you after that, this helps build up training examples for the topic to improve accuracy.
* **Suggested examples** - Aide automatically scans new conversations for examples for existing topics and adds them as suggestions for you to approve or rejected.
* **Outlier examples** - Aide automatically reviews examples present in every topic, even if they were manually added by your team, to ensure that they match the topic name, description, and other examples. In the following example, the message is asking for an update on their order, but as noted above, the topic we are looking at is “Book views”, so Aide is suggesting that it be moved to the correct topic, “**Order tracking**” and is presenting a button for you to do click and it will do so. You also choose keep it where it is by clicking “keep”.
An example that does not belong to “Book reviews”, Aide wants to move it to the “Order tracking” topic.
More importantly, when new examples are added, Aide compares them with the existing examples to determine if they really belong in that topic. If they don’t, Aide will suggest moving it to a more fitting topic. This is to ensure the consistency training examples and maintain the topic integrity for precise and accurate detection.
## **Managing existing examples**
For a given example, you can either switch it from positive to negative (or vice versa), move it to a different topic, or delete it entirely.
# Custom integration
Source: https://docs.aide.app/developers/custom-integration
Connecting a custom integration to Aide
### Custom integrations
This document provides a simple OpenAPI specification for the minimum API requirements needed to connect your data systems with Aide. We maintain flexibility regarding the fields returned in responses, requiring only a server endpoint with a URL and API token that implements three core endpoints: listing latest orders, retrieving individual order details, and fetching customer information.
````openapi: 3.0.0
info:
title: Order & Customer Management Integration API
version: '1.0.0'
description: |
Specification for integrating order and customer management platforms with Aide AI.
This API defines the minimum required endpoints and data structures
while allowing flexibility in implementation details.
Aide AI will poll these endpoints regularly (approximately once per minute)
to sync order and customer data.
servers:
- url: https://api.{platform}.com/v1
variables:
platform:
default: yourplatform
description: Your platform's API domain
paths:
/latest_orders:
get:
summary: List most recently updated orders
description: |
Retrieve the most recently updated orders, sorted by last update time in descending order.
This endpoint will be polled approximately once per minute by Aide AI to detect changes.
Implementation must provide at least the order ID and last update time.
parameters: - in: query
name: page_size
schema:
type: integer
minimum: 1
maximum: 100
default: 50
required: false
description: Number of orders to return per page - in: query
name: page_token
schema:
type: string
required: false
description: Token for pagination. Can be implemented as cursor or page number
responses:
'200':
description: Successfully retrieved orders
content:
application/json:
schema:
type: object
required: - orders
properties:
orders:
type: array
items:
$ref: '#/components/schemas/OrderSummary'
next_page_token:
type: string
description: Token for retrieving the next page of results
/orders/{orderId}:
get:
summary: Get order details
description: Retrieve detailed information about a specific order
parameters: - name: orderId
in: path
required: true
schema:
type: string
responses:
'200':
description: Successfully retrieved order details
content:
application/json:
schema:
$ref: '#/components/schemas/OrderDetail'
'404':
description: Order not found
/customers/{customerId}:
get:
summary: Get customer details
description: Retrieve detailed information about a specific customer
parameters: - name: customerId
in: path
required: true
schema:
type: string
responses:
'200':
description: Successfully retrieved customer details
content:
application/json:
schema:
$ref: '#/components/schemas/CustomerDetail'
'404':
description: Customer not found
components:
schemas:
OrderSummary:
type: object
required: - id - updated_at
properties:
id:
type: string
description: Unique identifier for the order
updated_at:
type: string
format: date-time
description: Time the order was last updated # Additional optional fields can be included
OrderDetail:
type: object
required:
- id
- created_at
- total_amount
properties:
id:
type: string
description: Unique identifier for the order
display_id:
type: string
description: Human-readable order identifier (e.g., order number, confirmation number)
created_at:
type: string
format: date-time
description: Time the order was placed
updated_at:
type: string
format: date-time
description: Time the order was last updated
total_amount:
type: number
format: float
description: Total amount for the order
currency:
type: string
description: Currency code (e.g., USD, EUR)
canceled_at:
type: string
format: date-time
nullable: true
description: Time the order was canceled, if applicable
customer_id:
type: string
description: ID of the customer who placed the order
status:
type: string
description: Current status of the order
billing_address:
type: object
description: |
Billing address as a JSON object. Common fields include:
- address1: Primary street address
- address2: Secondary street address (apt, suite, etc)
- city: City name
- province/state: Province or state
- zip/postal_code: ZIP or postal code
- country: Country code (e.g., US, GB)
You may include additional fields as needed for your platform
# Additional fields can be included
CustomerDetail:
type: object
required:
- id
properties:
id:
type: string
description: Unique identifier for the customer
first_name:
type: string
last_name:
type: string
email:
type: string
format: email
phone:
type: string
country:
type: string
description: Country code (e.g., US, GB)
billing_address:
type: object
description: |
Billing address as a JSON object. Common fields include:
- address1: Primary street address
- address2: Secondary street address (apt, suite, etc)
- city: City name
- province/state: Province or state
- zip/postal_code: ZIP or postal code
- country: Country code (e.g., US, GB)
You may include additional fields as needed for your platform
# Additional fields can be included
securitySchemes:
ApiKeyAut```
---
````
# Introduction
Source: https://docs.aide.app/index
Welcome to the Aide documentation
These docs detail the features offered in the dashboard and how to use them. For more advanced tips and recommendations, reach out to the Aide team for guidance specific to your use-case or industry.
**Aide features that work out the box:**
* **One click integrations**- for popular support and eCommerce and CRM platforms — Zendesk, Gorgias, Front, Shopify, WooCommerce, and Salesforce, with the ability to integrate with custom APIs as needed. See the next page for information on integrations.
* **Customer identification** - automatic retrieval of customer information from your CRM or eCommerce platform based on their email, phone number, order ID, or any other unique identifier.
* **Knowledge based drafts** - based on previous conversations, existing macros and canned templates, and your help center, enriched with real customer data from your CRM or eCommerce platform.
* **Topic suggestions** - Aide suggests a set of topics to help you get started. Topics categorize incoming messages for reporting and creating action-based automated scenarios as opposed to relying solely on your help center and response templates.
For use cases where customers want to simply get AI drafts, they can rely on knowledge articles.
For cases where accuracy is paramount and automation pathways and workflows can be defined, a structured approach using topics and scenarios is recommended.
# Integrations
Source: https://docs.aide.app/integrations/index
Connect Aide to your systems
Aide has several integrations. They are all one-click and connect in under 5 minutes. Zero engineering required.
### One-click integrations
* **Front** - for information about Aide’s Front integration, go [here](https://front.com/integrations/aide).
* **Zendesk** - for information about Aide’s Zendesk integration, go [here](https://www.zendesk.com/marketplace/apps/support/421826/aide-ai-support-automation/).
* **Gorgias** - for information about Aide’s Gorgias integration, go [here](https://www.gorgias.com/apps/aide-support).
* **Shopify** - for information on Aide’s Shopify integration, go [here](https://apps.shopify.com/aide).
* **Wooocommerce** - Aide’s WooCommerce is identical to Shopify integration.
* **Salesforce** - Aide’s Salesforce integration is documented separately.
### Custom integrations
# [See docs for integrating Aide here](/developers/custom-integration)
***
# Agent Panel
Source: https://docs.aide.app/products/agent-panel
Accelerate your agents with real-time AI
The Agent Panel is a tool helpful for teams that use Aide with human agents. It is available on Zendesk and Front. It presents real-time information for every conversation or ticket you view. It is composed of 4 sections:
1. **Context** - if the customer’s email address, phone number or order number is has an existing order or CRM record, it is retrieved and displayed as the first thing.
2. **Topic** - shows what topic was detected (if any). You can provide feedback with the thumbs up or down buttons. If no topics were detected, you can select a topic from the dropdown menu then click the checkmark button that appears.
3. **Generated draft** - if you’re using create draft in a Scenario that applied, you will find a pre-generated draft. If any knowledge articles were used they would be cited. To use the draft, click the Insert button, which will insert it into your Zendesk or Front composer. You can also provide feedback on the draft using the thumbs up or down buttons.
4. **Macros or suggested responses** - if there any Scenarios with suggested macros or responses, they would appear here.
As a last resort, potentially relevant macros or actions are listed towards the bottom.
Aide Agent Panel open in Zendesk.
Aide Agent Panel open in Front.
# AI Agent
Source: https://docs.aide.app/products/ai-agent
Automatically respond to customers and perform actions using AI
Aide's AI agent works behind the scenes to automatically handle customer conversations and perform actions based on your configured Scenarios. The AI agent can respond to customers using generative AI and automate various help desk and business operations.