Contact person: Business Client

Phone:Show

Email:Show

Location: Jerusalem, Israel

Budget: Recommended by industry experts

Time to start: As soon as possible

Project description:
"I’m ready to turn a proof-of-concept into a working SaaS that lets contractors drop a set of drawings into the browser and walk away with clean, structured data. You’ll own end-to-end development and help me ship an intuitive product that feels native on Web, Mobile, and Desktop.

Core workflow
• File intake – users upload PDFs or image plans; the service auto-detects orientation, page order, and scales.
• AI takeoff engine – first milestone focuses on accurate surface-area calculations (edges, cut-outs, slabs and other metrics can follow).
• Drawing outputs – generate 2D views, lightweight 3D previews, and a unit-by-unit summary for multi-family projects.
• Clarification assistant – flag missing notes, dimensions, or mismatched scales and update reports as the user resolves issues.
• Confidence scoring – surface low-certainty areas so estimators can double-check before exporting.
• Exports – JSON, CSV, and ready-to-share PDF reports.
• Account layer – simple email / password auth today, with room to bolt on OAuth or SSO later.
• Billing – monthly plans with a trial gate; I’ll handle Stripe keys.

Tech stack I’m leaning toward
• React / [login to view URL] front end
• Supabase (or Firebase) for auth, storage, and row-level security
• Node-based API for orchestration
• Poppler / pdf-lib plus OpenCV, Tesseract or TrOCR for parsing
• OpenAI (or Claude / Gemini) for LLM prompts and error-flag reasoning
• Optional: CAD libraries if it speeds 3D previews

What I need from you
1. Set up a clean repo, CI, and infrastructure-as-code (Vercel or similar).
2. Build the upload → AI takeoff → report pipeline with logging and retry logic.
3. Implement the first “surface area” model and expose it behind REST or GraphQL.
4. Deliver a responsive dashboard and a minimal billing gate.
5. Document deployment so I can spin up staging environments on demand.

Nice-to-have experience
• Geometry kernels or CAD SDKs
• Prior SaaS launch cycles (pricing tiers, usage metering)
• YOLO or custom vision models for symbol detection

Acceptance
A live staging URL and repo walkthrough demonstrating:
✓ Upload of at least three distinct plan sets
✓ Surface-area totals matching a provided ground-truth spreadsheet within ±3 %
✓ JSON / CSV / PDF export, plus basic email/password sign-up

If you love building with AI, enjoy clean code, and can communicate progress every few days, let’s chat!" (client-provided description)