PortfolioCase study
BV-01Live prototype · mock data · team
Bhakti Vilas
Devotion as behavioural health.
An elder-focused wellness prototype for India built around bhajan — the health meaning is coded into devotion, not front-loaded like a clinical app.
5 of 8
commits by Tushar in a team build
Structural
3 by Shivali, all on 1 Aug 2026
11
sources behind the answer to a mentor’s “why bhakti?”
Structural
Evidence keyMeasuredStructural
The problem
Elder care has no shortage of products — it has a shortage of clarity.
No trusted, verifiable layer tells the family who pays whether their parent is getting safe, dignified care. Tushar’s refined hypothesis: the family ends up as the system integrator, despite being the least present.
The market doesn’t appear to have a shortage of products. It appears to have a shortage of clarity.
Research (team)
Distance was never the variable. Availability was.
Team research from Prashant’s pod — family surveys (n=23 deep-dive, n=47 broad) and a healthcare-worker survey (n=12). Tushar’s part was segmentation, refined hypotheses and interview guides.
“Distance was never the variable. Availability was.”
— Seven-day series, Day 3 “Insure the visit, don’t vet the person.”
— Refined hypothesis — proposed, never built
Insight Financial burden ranked dead last as a challenge: 1 of 70 mentions in the team’s interviews.
The bet: Madhu Mukti
The health meaning is coded, not front-loaded.
- Could have
- A clinical app that names the condition upfront
- Chose
- Devotion first: bhajan, temple walks, WhatsApp messages
- Because
- The team’s research found people more willing to attend a kirtan than a diabetes screening.
- Could have
- A funnel that pushes everyone to screening
- Chose
- A staged reveal where staying devotional is success
- Because
- The funnel narrows in trust and consent, not effort; its shares are directional estimates, not data.
The prototype
A working prototype, live — not a deck.
Four static pages on mock data: no backend, no database, no AI. The session-booking flow was verified end to end, with a mobile audit at 390 and 360 px.
- Learn
- Pick a date
- Circle on a map
- Payment
- QR pass
Evidence
Defended with research, not with usage.
The mentor’s three questions were answered in writing — and the answer names its own limit: “a well-evidenced hypothesis with real mechanisms, not a proven guarantee.”
0
console errors in the 390 and 360 px mobile audit
Measured
What isn’t measured
- No users and no product metrics; the survey numbers describe the market, not this prototype.
- Translation covers about 90 of ~500 strings, and the medical copy is unreviewed.
- The team PRD carries no author names.
What I learned
Honest framing is part of the design.
Disclose the gaps
Partial translation and unreviewed medical copy belong on the page, not under it.
Never dress devotion as clinic
Devotional-health framing is not clinically vetted, and shouldn’t read as if it were.
Credit the team exactly
Five of eight commits, and a team PRD without names: say both.
Sources
Everything here is backed by a real artifact.
- Research
- PRD
- Build ledger
- Feedback
- Readme