/ 01 — Mission
Tell people why a locality is a smart choice, not just where to look.
Ooru — Kannada for one's own place, one's town — is the AI locality-intelligence engine for Bangalore house hunters. It combines real air quality, verified school proximity, and metro access into a reasoned advisory a person can trust and revisit. ooru.maahaa.dev
/ 02 — The problem
What we actually solve
Rental discovery in Bangalore is solved by maps and listings. bengaluru.rent shows real crowdsourced rents on a Google Map with zero friction. 99acres and Magicbricks show volume. None of them answer the decision question a family asks: is this a good place to live, and why?
That question needs quantitative livability — clean air, good schools nearby, a workable commute — turned into a clear, reasoned recommendation. That is the gap. That is the only thing Ooru needs to be excellent at.
/ 03 — The one rule
Honesty comes before features
The advisory must only state findings the pipeline actually computed. Right now it does not — and that is the single most important thing to fix.
The report claims "ICSE schools nearby" and "metro under 3km", but the school and metro nodes
return {} — that data never reaches the LLM. The claims are hardcoded boilerplate. Until this is
fixed, Ooru is less trustworthy than a competitor with no analysis at all, because it asserts
findings it never made.
/ 04 — Where we win
Reasoning and memory, not the map
Not on the map. bengaluru.rent has a multi-year crowdsourced rent-data lead and a frictionless, no-login model. Racing them on pins and matching is a losing fight.
We win on the two things they structurally lack: verified multi-factor reasoning (AQI numbers, school distance, metro access, fused into a judgement) and authenticated personal memory (saved searches, history, side-by-side comparison). Their strength is data volume. Ours is intelligence and trust.
Map tools show you where people rent. Ooru tells you why a locality is — or isn't — a smart choice.
/ 05 — Success criteria
What good looks like
- Every advisory claim is traceable to real pipeline output. No fabricated findings, ever.
- Schools and metro become honest differentiators — real counts and distances, surfaced in the report and persisted.
- The user can revisit and compare saved localities. Memory is the moat the anonymous competitor can't build.
- Recommendations within the existing stack. Keep the orchestration brain; add capability, don't rewrite blindly.
- Honesty over flash. Restore trust before shipping any new surface.
/ 06 — Read next