Side Project2026
Kindred Stays
Share more than a place.
Existing platforms (99acres, NoBroker, Facebook groups) surface what people can afford — not how they actually live, so incompatibility surfaces weeks after move-in, when exiting costs a lost deposit and days of effort. Kindred Stays matches on living habits and dealbreakers, predicts conflicts in plain language, and builds India-specific trust.
Solo PM & builder — discovery → spec → no-code MVP
- >60%
- Activation target (full profile)
- >40%
- Match → meetup
- <48h
- Time to first match
- 7
- Competitors torn down
What I actually did.
- Validated with JTBD mapping, three MOM-test personas, and a 7-player competitive teardown — every finding grounded in past behavior, not hypothetical preference.
- Scope discipline: cut onboarding 10 → 5 fields after observing drop-off at fields 7–8, before a line of the app was built.
- Spec-first: an explicit compatibility-engine design, a RICE-ranked 30+ backlog, go/no-go bars, and written stack rationale (Bubble · Supabase · Twilio).
- A go/no-go metrics framework defined up front — honest guardrails governing decisions, not claimed traction.
How it works.
1
Weighted Compatibility Scoring
Asymmetric dealbreaker vs. nice-to-have weighting, reported 0–100% with full per-attribute traceability.
2
Conflict Prediction
Not a bare score — a human-readable reason: "You keep a strictly vegetarian kitchen; they cook meat daily."
3
India-Specific Trust
Aadhaar-backed verification, rich profiles (photos, video, reviews), and anti-broker design.
The deck.
Open full screenSkills & tools
Product DiscoveryJTBDRICE PrioritizationMarketplace0→1No-code MVP