Getting Started¶
This section helps you quickly understand how to use the Madevo API and how the core pieces fit together.
How the API works¶
The Madevo API is built around a few core concepts:
- Authenticate
- Upload and define data (datasources)
- Query or analyse data (assistant)
- Automate workflows (agents)
- Receive real-time updates (WebSocket)
Basic flow¶
A typical integration follows this flow:
1. Authenticate¶
Start by logging in to obtain a JWT:
You will use this token for all secure endpoints.
2. Upload data¶
Upload a CSV or ZIP file:
This returns: - a storage filename - detected columns - a sample of your data
3. Create a datasource¶
Define how your data should be interpreted:
This step: - stores metadata - defines schema (timefield, metafield, values) - triggers background data import
4. Insert or query data¶
You can either:
- Insert rows directly:
- Or query data:
5. Use the assistant¶
Create a conversation:
Send a message:
The assistant will: - analyse your data - generate insights - optionally return queries and visualizations
6. Automate with agents¶
Create an agent task:
Run it manually or schedule it:
7. Listen for real-time updates¶
Connect to the WebSocket:
Receive events for: - assistant streaming - datasource processing - agent execution
Key components¶
Authentication¶
Handles login, token refresh, and session management.
Datasources¶
Where your data is stored and structured.
Assistant¶
Natural language interface for querying and analysing data.
Agents¶
Automated workflows that run tasks using your data.
WebSocket¶
Real-time event stream for async operations.
What to read next¶
- Authentication to understand how to securely access the API
- Response Format to understand how responses are structured
- Pagination to handle large datasets ```