AVENUOR — Home Asset Management SaaS
AVENUOR is a home and society asset management SaaS live at avenuor.com. It gives homeowners a central registry for appliances, vehicles and building assets — with AI-powered invoice extraction, per-asset health scoring and warranty reminders — and gives housing societies a multi-user console for shared registers, work orders and compliance documents. Built with Next.js and React, deployed on Vercel.
- Role
- Full-stack development.
- Year
- 2026
- Stack
- Next.js, React, AI Integration, SaaS, Vercel
The problem
Household and society asset records live in drawers and spreadsheets: warranties expire unnoticed, service history disappears with the last committee, and a repair-or-replace decision gets made with no data. Homeowners needed effortless capture — photograph an invoice, done — and societies needed shared registers with roles, work orders and an audit trail.
How it was built
01An asset registry that starts from an invoice
The unit of the product is the asset, but the unit of user effort is the invoice. Users photograph or forward a purchase invoice and an AI extraction step turns it into structured data — product, price, date, vendor — validated against a schema before anything is written. Manual entry exists, but the design assumption is that nobody maintains a registry they have to type into.
02Health scoring and repair-vs-replace
Each asset carries a health score derived from its age, service history and expected lifetime, surfaced alongside a repair-versus-replace view. The point is not a prediction dressed as certainty — it is turning 'the lift AMC lapsed and the pump is nine years old' from tribal knowledge into something a committee can see and budget against.
03A society console with real roles
Housing societies get a multi-user workspace: shared asset registers across buildings, work orders with photos and bills attached, a compliance vault for certificates, and multi-year capital planning. Committee roles are enforced server-side — treasurer, secretary and member see and do different things, and every change records its actor.
04Freemium gating done server-side
The free tier covers the first ten assets; premium unlocks beyond that. Plan limits are enforced at the API, not hidden in the UI — client-side gating is a suggestion, and a pricing model only means something if the server agrees with it.
Stack decisions
Why each piece was chosen, rather than just what was used.
Next.js + React
Server components keep the AI extraction calls and their credentials server-side, and a single deployable unit keeps the operational surface small for a lean product. Marketing pages and the app share one codebase and one deploy.
AI extraction with structured outputs
Invoice parsing is a judgement task rules do badly — layouts vary endlessly. A model constrained to a schema, with validation and retry on mismatch, turns arbitrary invoice photos into rows the rest of the product can trust.
Vercel (Mumbai region)
The user base is Indian households and societies; serving from the bom1 region keeps latency low where the users actually are, with preview deployments for every change.
What it does
- Central asset registry for appliances, vehicles and building assets
- AI invoice extraction from photo upload or email forwarding
- Per-asset health scoring with repair-vs-replace analysis
- Warranty and service schedule tracking with reminders
- Digital document vault
- Society console: shared registers, work orders, compliance vault, capital planning
- Freemium model — free for the first 10 assets, premium at ₹99/month
Questions this raises
How does AI invoice extraction actually work?
The invoice image goes to a vision-capable model with a strict output schema — product, price, date, vendor — and the response is validated before it touches the database; a mismatch retries rather than storing garbage. The model handles layout variety, the schema handles trust. Users confirm the extraction, so the AI proposes and the human commits.
Why build asset management as a SaaS rather than an app with local storage?
Because the hard requirements are shared state and continuity: a society committee changes every year, and a household's records need to outlive any single phone. Server-held data with roles and an audit trail is the product; local storage would just be a nicer drawer.
How do you enforce a freemium limit properly?
At the API. The server counts assets against the plan on every create, and the UI merely reflects what the server will allow. Client-side gating gets bypassed by anyone with DevTools open — and on a paid product, that is revenue leaking through presentation logic.
Related work
Building something similar?
Naman Gundaniya takes on full stack and AI projects like this one. Available for hire, replies within 24 hours.