> Hi HN, I'm Martin, a solo developer from Germany. > > I've been house-hunting for the past year, and the biggest pain point was always the same: a listing says "oil heating, built 1992, energy class G" — but what does that actually cost to fix? > > I kept building spreadsheets for every property (roof replacement? windows? heat pump? which government subsidies apply?) and eventually thought: this should be automated. > > So I built Immo-Lens as a side project. You paste a link from any German property portal (ImmoScout24, Immowelt, etc.) and it returns within ~30 seconds: > > - Renovation cost estimates in 3 scenarios (minimum / standard / full EH55 retrofit) > - Matching government subsidy programs (Germany has KfW low-interest loans up to €150k for energy-efficient renovations) > - Risk analysis (what might be more expensive than expected) > - Total investment overview including transaction costs > > *Tech stack:* React 19 + Vite frontend, Node.js/Express backend, Google Gemini for listing analysis, PostgreSQL, Cloudflare R2 for storage, deployed on Railway. > > The AI extracts structured data from listing text (build year, heating type, energy certificate, roof condition, etc.) and feeds it into a deterministic cost model with current German construction prices. Subsidy matching is rule-based, not AI — the funding programs have strict eligibility criteria that need to be reliable. > > *Important caveat:* This doesn't replace a proper building survey. It's more like a better calculator — it helps you ask the right questions before spending €500k+. You should still bring a surveyor to the viewing. > > The basic analysis is free and works without signup. Power users who want to save projects, upload photos, and add notes can create an account. There's a small paid tier to cover hosting and API costs — no VC funding, just a side project that needs to sustain itself. > > Would love feedback, especially from anyone familiar with European PropTech or construction cost estimation. Are there similar tools in other markets? Happy to discuss the technical approach.