Smart, private, AI-powered wellness for Punjab's youth — a fitness, productivity, and mental-wellbeing platform that already runs, live, on a server budget any university can afford.
Most digital wellness apps solve one slice of this — fitness, or mood, or focus — never all three. The AI-powered ones that try usually route every message through an expensive external LLM, adding cost, latency, and a data-privacy question nobody asked for.
A generation that is simultaneously underweight and overweight can't be served by one generic plan — it needs personalised guidance. And it needs that guidance running on infrastructure Punjab's universities and Business Incubation Centers can actually afford, not a cloud AI bill scaled for Silicon Valley.
FitAI is already live at mdsalar.site/fitness — a finished PHP 8.1 MVC application with 20+ controllers, 30+ database tables, 40+ working screens, and a packaged Android app. Not a mockup, not a slide-deck idea.
It isn't the chatbot — it's what's missing. No OpenAI or Claude API bill per query. No internet dependency for the core engine. No user conversation leaving the server. Rule-based intelligence, chosen deliberately for cost, speed, privacy, and reliability.
Five inputs become four personalised outputs, with no outbound API call in between. The AI Coach chatbot uses the same philosophy: a rule-based heuristic engine (FitnessCoach::respond()) that retains the last 20 messages in-session for context, without ever leaving the server.
Flip through the entire product: landing page, onboarding, AI daily plans, workout logging, habits, timetables, journaling, achievements, Monk Mode, and profile — captured directly from the live deployment.
Already deployed and in active use. Because the product exists rather than existing only as a pitch, FitAI carries materially lower implementation risk than a typical early-stage proposal.
| Criterion | Wt. | How FitAI Delivers |
|---|---|---|
| Innovation & Originality | 20% | A privacy-first, rule-based AI engine generating plans, coaching & challenges instantly — no per-query cost, no external LLM dependency. |
| Public Benefit | 20% | Improves physical health, mental focus & productivity for students, on hosting cheap enough for genuinely wide access. |
| Social Impact | 15% | Friends/chat system drives peer accountability; journaling & Monk Mode support youth mental wellness and digital-wellness habits. |
| Feasibility | 15% | Already functioning — working MVC architecture, secure auth, dual-database layer, live today. |
| Scalability | 10% | Lightweight PHP/MVC + shared hosting means cheap replication across universities & TEVTA institutes province-wide. |
| Economic Viability | 10% | No LLM API billing supports a sustainable freemium model, plus B2B licensing to universities, gyms & corporates. |
| Sustainability | 5% | A local, rule-based engine means no recurring third-party AI costs and no dependency on external providers. |
| Presentation Readiness | 5% | Working architecture and a phased rollout plan demonstrate implementation readiness today. |
FitAI is live, tested, and ready for pilot deployment through BGNU's Business Incubation Center today.