Introduction
Welcome to HappyLoop, the AI Analyst that connects to your SaaS APIs and transforms raw data into actionable insights.
This guide walks you through the initial setup process: connecting your APIs, defining schemas, and configuring operations. These steps are typically performed by a developer or someone with technical API access.
After following these steps, your AI Analyst will be ready to use. You will be able to ask questions in plain English and get instant answers from your live data.
In this article
Login
Go to app.happyloop.com and log in with your business email. Your account has already been set up by our technical team.
Navigation: Sidebar Overview
Use the left-hand sidebar in the HappyLoop interface to access the key setup areas:
- Organizations
- Definitions
- Schemas
- Operations
- Test Area
Each section below corresponds to one of these menus.
1. Organization
An Organization represents your company within HappyLoop. It acts as the top-level container for:
- API integrations
- Schema and operation definitions
- User access
Create or select your Organization to begin the setup process. Each Integration is linked to an Organization.
2. Definitions (API Setup)
In the Definitions section, configure the details of your external API:
- Base URL
- Authentication credentials (e.g., API keys, bearer tokens)
- Environment (e.g., production, staging)
- Metadata: API name, short description, logo
This step sets the foundation for all subsequent configurations. Unlock π the Integration to edit and save changes.
3. Schemas (Data Structures)
Schemas define the structure of your APIβs inputs and outputs. They help HappyLoop understand and model the shape of your data so it can answer accurately.
Creating or Editing a Schema
- Use + Create New Schema to define a new structure
- Unlock π an existing schema to edit sand save
Schemas are made up of fields, each representing a specific data point.
Edit Field Form (Field Properties)
Each field includes the following:
-
Name β Required. Must match the API parameter (e.g.
paymentDate
) - Description β Short explanation (e.g. βOrder payment dateβ)
- Location β Where the field appears: Path, Query, Header, Body
- Type β Data type (e.g. Datetime, String, Integer, List)
- Default Value β Optional fallback value
- Examples (optional) β Illustrative sample values
-
Optional β Allows
null
as a value - Skip β Prevents AI from autofilling; uses default instead
Schemas can be reused and composed to reflect complex API structures.
4. Operations (Endpoint Mapping)
Each Operation represents a single interaction with your API β usually a data fetch or update.
Define an Operation
-
Name β Unique label (e.g.
GetOrders
) - Type β e.g. GET, POST, PUT, DELETE
- Description β What the endpoint does
- Dictionary Tags β Keywords for triggering via chat
-
Endpoint Configuration β HTTP method and path (e.g.
/orders/client/{client_name}
) - Input Schema β Data required for the request
- Output Schema β Expected structure of the response
- Count / Offset Fields (optional) β For paginated endpoints
- Code Section β Add lightweight transformation logic to extract relevant response data
Use operations to define how HappyLoop communicates with each endpoint in your API.
5. Test Area (Validation & Debugging)
Use the Test Area to validate your setup and preview API responses.
Key Functions
- Send sample requests
- View raw API responses
- Generate transformation logic via Suggest Code
β οΈ Use only sandbox or dummy data when using Suggest Code. Input in this section may be shared with AI models to assist code generation.
Note:
- Large responses may limit code suggestion capability
- Outside of this area, no data is shared with external models
The Test Area helps refine your logic and ensure everything is working correctly before use.
Use the Analyst
Youβve now:
- Set up your organization
- Connected your API
- Defined schemas and operations
- Validated everything via test calls
HappyLoop is now ready to analyze your data. Go to Chat and start asking questions.
Comments
0 comments
Please sign in to leave a comment.