Skip to Content

Document Q&A

Build a system that answers questions from your documents.

Overview

Upload PDFs, Word docs, and text files, then ask questions in natural language.

Time to deploy: ~10 minutes

Quick Deploy

git clone https://github.com/xenoISA/quickstart-document-qa
cd quickstart-document-qa
pip install -r requirements.txt
cp .env.example .env
python main.py

Features

  • PDF, DOCX, TXT, MD support
  • Semantic search with citations
  • Conversation memory
  • File upload API

Implementation

from isa_agent_sdk import Agent
from isa_agent_sdk.rag import DocumentStore
 
store = DocumentStore(collection="documents")
 
agent = Agent(
    name="doc-qa",
    model="claude-sonnet-4-20250514",
    tools=[store.as_tool()],
    system_prompt="Answer questions based on the uploaded documents. Always cite your sources."
)
 
# Upload endpoint
@app.post("/upload")
async def upload(file: UploadFile):
    await store.ingest_file(file)
    return {"status": "indexed"}
 
# Query endpoint
@app.post("/query")
async def query(question: str):
    response = await agent.run(question)
    return {"answer": response.content}
Was this page helpful?
Home