Docs Agent

Why deploy Docs API Agent on Render?

Docs API Agent is an AI-powered documentation chatbot that uses RAG (Retrieval-Augmented Generation) to let users query product documentation through natural language. It solves the problem of users having to manually search through documentation by combining Claude AI with vector search to retrieve and synthesize relevant answers from your docs.

This template wires together a FastAPI backend, React frontend, and PostgreSQL with pgvector already configured—including automatic database migrations, vector index creation, and service-to-service communication. Instead of manually setting up pgvector extensions, configuring embedding pipelines, and debugging connection strings, you get a working RAG documentation agent with one click. Render's Blueprint deployment handles the orchestration, so you just add your API keys and the entire stack deploys with proper health checks and environment variable linking between services.

Architecture

What you can build

After deploying, you'll have a documentation chat agent that lets users ask questions about your product docs in natural language and get answers with source citations. The system uses vector search to find relevant documentation chunks and Claude to generate responses. You can upload Markdown files or OpenAPI specs through the web UI and immediately start querying them.

Key features

• HNSW Vector Indexing: Uses pgvector with HNSW indexing for 10-100x faster semantic search over documentation embeddings.
• Multi-format Document Parsing: Supports uploading and processing both Markdown files and OpenAPI specifications for RAG ingestion.
• Persistent Conversation Context: Maintains multi-turn conversation history with context awareness for coherent follow-up responses.
• Source Citation Display: Returns which specific documents informed each AI response along with real-time confidence scores.
• One-click Render Deployment: Pre-configured render.yaml blueprint handles database migrations, vector index creation, and service linking automatically.

Use cases

• Support engineer adds internal docs so teammates can self-serve answers
• Developer uploads API specs to help users troubleshoot integration issues
• Technical writer tests if documentation answers common customer questions
• Startup founder creates a chatbot for their product's knowledge base

What's included

Prerequisites

• Anthropic API Key: API key for Claude AI, which powers the conversational responses in the documentation agent.
• OpenAI API Key: API key for OpenAI, used to generate vector embeddings for semantic document search.

Next steps

1. Open the frontend URL and upload a Markdown file from the test-docs folder — You should see the document appear in the document list and show a 'processed' status within 30 seconds
2. Test the chat by asking a question about the documentation you just uploaded — You should receive an AI-generated response with source citations and a confidence score displayed below the answer
3. Configure your own documentation by uploading an OpenAPI spec file through the drag-and-drop interface — You should see the API endpoints parsed and indexed, then be able to ask questions like 'What endpoints are available?' and get accurate responses