Files · AWS Bedrock Contract Clause Extraction for SMB Legal

# AWS Bedrock Contract Clause Extraction for SMB Legal
 
> Automatically extract key clauses, dates, and parties from contracts using AWS Textract and Bedrock, with structured output and cost tracking.
 
A tutorialized reference solution from [reaatech.com](https://reaatech.com), demonstrating how to build production-grade AI systems with the `@reaatech/*` package family.
 
## Architecture
 
```
PDF Upload → unpdf (pdf.js) → Textract (OCR fallback) → Bedrock (Claude Sonnet 4)
→ REAA doc-extraction → structured-repair → PostgreSQL (Drizzle ORM)
```
 
## Setup
 
```bash
# Install dependencies
pnpm install
 
# Copy env vars and fill in your credentials
cp .env.example .env
 
# Push the database schema
pnpm drizzle-kit push
 
# Start the dev server
pnpm dev
```
 
## API
 
### Extract clauses from a PDF
 
```bash
curl -X POST http://localhost:3000/api/extract \
  -F "file=@contract.pdf"
```
 
### Extract clauses from plain text
 
```bash
curl -X POST http://localhost:3000/api/extract \
  -H "Content-Type: application/json" \
  -d '{"text": "This Agreement is made between..."}'
```
 
### List contracts
 
```bash
curl http://localhost:3000/api/contracts?limit=10&offset=0
```
 
### Get a single contract
 
```bash
curl http://localhost:3000/api/contracts/<id>
```
 
### Create a contract manually
 
```bash
curl -X POST http://localhost:3000/api/contracts \
  -H "Content-Type: application/json" \
  -d '{"filename": "example.pdf"}'
```
 
## Project layout
 
```
app/                  Next.js App Router pages + API routes
src/                  services, lib, adapters
tests/                vitest suite (mirrors src/)
packages/             API references for every dependency (read these first)
DEV_PLAN.md           build plan for this recipe
```
 
## License
 
MIT — see [LICENSE](./LICENSE).