You are my product co-founder. Help me sharpen the vision for a 60-minute hackathon build.
PROJECT: RFPilot AI — an AI proposal intelligence platform that ingests RFPs, grants, and tenders, extracts compliance requirements, auto-drafts proposal sections, scores readiness, and tracks submission
workflows.
THE PROBLEM (real, painful, validated)
Mid-market firms and nonprofits chase RFPs and grants worth billions, but the proposal process is brutal:
1. Reading a 60-page RFP and pulling out every "must" / "shall" requirement takes 3–5 hours
2. Drafting the executive summary, project narrative, and compliance matrix takes days
3. Missing one mandatory clause = automatic disqualification
4. Teams scatter requirements across Word, Excel, and email — nothing is the source of truth
5. Deadlines slip because no one knows how "ready" the submission actually is
6. Win/loss patterns are invisible — same mistakes repeat across submissions
Existing tools (Loopio, Responsive, RFPIO) cost $20k–$80k/year, are enterprise-only, and weren't built when LLMs could do the actual writing. The gap is an AI-first, mid-market-priced proposal copilot.
TARGET USER PERSONA
"Sarah" — 34, Proposal Manager at a 120-person govtech firm. Tracks 8–12 active RFPs simultaneously. Currently uses Word + SharePoint + a homemade Excel compliance matrix. Loses ~15 hours per RFP to manual
requirement extraction. Has missed deadlines twice this year due to last-minute compliance gaps.
WHY NOW
- LLMs now reliably extract structured data from long unstructured docs
- Long-context windows make full RFP ingestion trivial
- Government and nonprofit RFP volume hit record highs in 2025
- Mid-market firms can't afford Loopio but desperately need its outcome
PRODUCTION OUTCOME I WILL SHIP IN 60 MINUTES
A deployed, login-gated web app at a public Lovable Hosting URL with:
- Clean dashboard with active RFPs, upcoming deadlines, readiness scores
- "New RFP" flow: paste RFP text → AI extracts requirements, deadline, eligibility, sections, scoring criteria in under 10 seconds
- Requirements matrix per RFP (must/shall/should) with assigned owner and status
- AI section drafter: generate executive summary, project narrative, technical approach, compliance matrix per RFP context — under 8 seconds per section
- Readiness score (0–100) auto-computed from requirements completion + draft coverage
- Kanban pipeline: Discovery → Drafting → Internal Review → Submitted → Won/Lost
- Analytics: win rate, average readiness at submission, top failure reasons
- Pre-seeded "Demo Mode" with 6 realistic RFPs, 80 extracted requirements, 4 generated drafts so judges experience the product instantly
WHY THIS BEATS OTHER CHALLENGE TOPICS
- Demoable in 90 seconds with an obvious "wow" moment (paste RFP → watch AI parse it live)
- Universal: every judge has seen or written a proposal
- Buildable in 60 min with Lovable + Cloud + AI Gateway without faking the core value
- Scoped: ingestion is text-paste for MVP (file upload mocked as text extraction), but data model is production-shaped — Day 61 we plug in real PDF parsing, not a rewrite
NON-GOALS (out of scope)
- Real PDF/DOCX file parsing (text-paste in MVP — we mock the OCR layer)
- Real DocuSign / submission portal integrations
- Real CRM (Salesforce/HubSpot) sync
- Multi-tenant team accounts (single-user workspace only)
- Native mobile app
SUCCESS CRITERIA AT MINUTE 60
✅ Public URL works on judge's laptop
✅ Sign-up → dashboard → "Load Demo Data" → demo flow runs without a single bug
✅ Paste-RFP → AI extracts requirements in under 10 seconds
✅ AI generates an executive summary draft in under 8 seconds
✅ Kanban drag-drop, readiness score updates live, analytics charts render
ELEVATOR PITCH
"RFPilot AI reads your RFP, extracts every requirement, drafts your proposal sections, scores your readiness, and tracks every submission — turning 40 hours of proposal work into 40 minutes. Built for Sarah, not
for the Fortune 500."
Lock this vision in. Every subsequent prompt should serve this exact frame — no scope creep, no generic project-management detours.
This project was built with Lovable.
Live app: https://proposal-pilot-ai-56.lovable.app
Continue developing this project in the Lovable editor.
- Ship faster: describe what you want to build and Lovable handles the code.
- Stay in sync: every change made in Lovable is committed straight to this repository.
- Full ownership: this code is yours. Push to
mainon GitHub and your changes sync back into Lovable, ready for your next prompt.
Prefer working locally? You need Node.js and npm — install with nvm.
git clone <this-repository-url>
cd <repository-name>
npm i
npm run dev