Tables & endpoints

In the previous lesson you implemented authentication in DYPAI. Now you design the core of your app—starting with tables and endpoints. If you are not sure what to build yet, you can plan with your AI agent: open a new chat and describe what you want (for example a task management app). You are only exploring recommendations at first—structure, style, and features—nothing is built yet.

What this video covers

From plan to implementation
Once you are happy with the plan, you ask the agent to start implementing it in DYPAI. It uses the MCP tools, creates the tables, and you can confirm them under Database in your project (for example a task table with the right columns).

Endpoints
The agent also creates REST endpoints and tests them. That is deliberately basic at this stage—simple flows like a start node plus a database node—because tutorials will grow in complexity later.

API view
In API, you see endpoints such as create task and list tasks. Opening one (for example list tasks) shows a minimal workflow: enough to prove the idea, not the final complexity.

When something fails
Errors can happen—that does not mean DYPAI is broken; it often means the configuration or generated flow needs a tweak. You tell the agent there is an error and ask it to fix it in DYPAI. It investigates, corrects, and you retry. In the video, a date field caused a save issue; after the fix, saves work and you can verify rows in the task table (including older test rows the AI created while iterating).

Iterate with new features
Want task groups? Ask the agent for that functionality—you do not need a different process. With the project already connected, you keep iterating: more tables, more endpoints, and small UI adjustments (for example adding a delete button, filters by group, a test group). The example UI stays simple on purpose; you could keep improving layout, sidebars, and pages—the point is speed and versatility: you focus on shaping the product while the agent helps build the backend.

Integrations
More integrations will keep appearing in the product over time.

Mindset
Keep the same loop: build, hit errors, feed them back to the AI, and let it repair the setup inside DYPAI.

What you'll learn

• Planning an app with the agent before touching implementation
• Letting the agent create tables and endpoints and test them
• Reading simple API workflows (start + database nodes)
• Debugging by describing errors to the agent and re-testing
• Verifying data in the Database UI
• Iterating (groups, filters, UI tweaks) without changing your overall workflow

Next

Continue to Storage for file attachments on tasks and how DYPAI handles buckets and the storage system.