Read this in: English · 简体中文
Upload a government or school procurement notice. AI tells you whether it's worth bidding.
The output is a one-page decision brief: Bid / Don't bid / Bid with caution, with a qualifications checklist, a budget attractiveness score, and a competitive density score.
# Install dependencies
pip install -r requirements.txt
# Configure API key
cp .env.example .env
# Edit .env, fill in DEEPSEEK_API_KEY (apply at https://platform.deepseek.com/)
# Run
uvicorn main:app --reload --port 8000Open http://localhost:8000 in a browser and drop in the notice file.
| Format | Notes |
|---|---|
| Digital PDFs are extracted directly; scanned PDFs use OCR (requires Tesseract) | |
| Word | .docx / .doc |
| HTML | Paste the notice page directly |
File size limit: 10 MB.
{
"verdict": "Bid with caution",
"verdict_reason": "Budget is reasonable, but technical specs are unusually specific",
"qualification_match": "Pass",
"qualifications": [
{ "requirement": "Business license (IT category)", "status": "pass" },
{ "requirement": "ISO 9001 certification", "status": "unknown" }
],
"budget_score": 4,
"budget_analysis": "500K RMB budget is above industry average",
"competition_score": 2,
"competition_analysis": "Specs name a specific brand and model, suggesting a preferred supplier may already be in mind",
"summary": "Budget is acceptable, but heavily customized technical specs hint at a likely incumbent. Evaluate carefully before bidding."
}# Windows
# Installer: https://github.com/UB-Mannheim/tesseract/wiki
# Check the Chinese Simplified language pack during install
# macOS
brew install tesseract tesseract-lang
# Linux
apt install tesseract-ocr tesseract-ocr-chi-simpytest tests/ -v- Backend: FastAPI + uvicorn
- PDF parsing: PyMuPDF + Tesseract OCR
- Word parsing: python-docx
- AI analysis: DeepSeek API (JSON mode)
- Frontend: Single-page HTML, no framework
| Variable | Notes |
|---|---|
DEEPSEEK_API_KEY |
DeepSeek API key (required) |
MIT