TenderFit scouts public GeM / BidPlus tenders for vehicle hiring (cab/taxi/SUV/MUV), collects bid documents and corrigenda, extracts eligibility and SLA constraints with citations, and scores Go/No-Go fit against a fleet operator profile.

This is not a single-agent summarizer. TenderFit uses a coordinated multi-agent workflow with verification and consensus:

• Scout (find relevant tenders)
• Collector (fetch docs + corrigenda with precedence)
• Extractor (structured extraction)
• Verifiers (independent citation checks)
• Arbiter + Fit Scorer (consensus + ranking + gap checklist)

• python3 -m tenderfit.cli scan --keywords "cabs taxi" --days 14 --top 30 --max-pages 20 --llm-filter
• python3 -m tenderfit.cli fetch --bid-id <ID> --out artifacts/<ID>/
• python3 -m tenderfit.cli evaluate --bid-id <ID> --company examples/company_profile.example.json --out reports/<ID>.md
• python3 -m tenderfit.cli shortlist --company examples/company_profile.example.json --top 10 --out shortlists/shortlist.csv
• Demo flow: python3 -m tenderfit.cli demo

The React UI mirrors the CLI demo flow with live pipeline progress, per-bid lanes, and shortlist insights.

• Start API: python3 -m tenderfit.web.server
• Start UI: cd tenderfit-ui && npm run dev
• Open: http://localhost:5173

We only fetch publicly accessible bid documents and do not require login.

1. Scout searches BidPlus listings (tokenized keywords + LLM filter).
2. Collector fetches bid docs and corrigenda with precedence.
3. Extractor pulls structured requirements with citations.
4. Verifiers independently check evidence + quotes.
5. Arbiter resolves conflicts and assigns fit scores.
6. Shortlist ranks bids and emits CSV.

• Python 3.11+, FastAPI (API + SSE), Pydantic
• OpenAI API (LLM filtering + evaluation)
• agent-browser (BidPlus navigation + document capture)
• React + Vite (Web UI)

1. Python 3.11+ and Node 18+
2. Create a virtualenv and install Python deps:
   • pip install -e .
3. Install agent-browser:
   • npm install -g agent-browser
   • agent-browser install
4. Set API key:
   • export OPENAI_API_KEY=...
5. Install UI deps:
   • cd tenderfit-ui && npm install

Run agent eval suites:

• python3 -m tenderfit.cli eval --suite quick
• python3 -m tenderfit.cli eval --suite scout
• python3 -m tenderfit.cli eval --suite collector
• python3 -m tenderfit.cli eval --suite extractor
• python3 -m tenderfit.cli eval --suite verifier
• python3 -m tenderfit.cli eval --suite arbiter
• python3 -m tenderfit.cli eval --suite shortlist See tenderfit/evals/README.md for details.

• OpenAI API (LLM scoring, filtering)
• agent-browser (vercel-labs/agent-browser)
• FastAPI + Uvicorn
• Pydantic
• React + Vite

Made with ❤️ by Priyank Mehta for AI Engineers Day Hackathon • Presented by OpenAI • Peak XV • Activate