../missions / resonate
Resonate
Universities have two states for a student — 'fine' and 'in crisis' — with nothing in between. Resonate is the daily practice built for the in-between, launching free at UF this fall.
- problem
- College mental health has a structural gap: universities built clinical supply (teletherapy, therapist networks) but nobody won daily student attention. Incumbents reach roughly 2–6% of eligible students; over half of students report high loneliness; and a large share with symptoms get no care at all. The students who need support most live between 'fine' and 'in crisis' — and have nothing they'd open every day.
- my_role
- Founder. I own the product, the code, the go-to-market, and the safety design — and I run the whole company through a multi-agent operating system I built, so strategy, research, red-teams, and build all compound.
- system_built
- A campus-native PWA built around one small daily loop and three deepening 'loops' — Solo (grounding), Social (anonymous, resonance-based connection), and Physical (the campus as a game board) — plus a therapist-ready Journey Map that turns Resonate into a warm referral into UF's Counseling & Wellness Center, never a walled garden.
- proof
- The v2 prototype is code-complete and in launch-gate hardening (Next.js · TypeScript · Supabase · UF Navigator LLM) with a green test suite. Safety is a 3-tier crisis exit that runs client-side and server-side and is always user-owned — detection supplements, never gates. Built UF-native: @ufl.edu auth, real campus buildings, CWC/GatorWell routing.
- proof_status
- in-progress build scope
- proof_note
- Product/build scope, not clinical or outcome validation. This is a free learning launch, not a finished product — everything gates on one honest question: do students voluntarily come back?
- artifact_note
- Private build artifact; add screenshots, demo, or repo link once the launch cohort is live.
- why_it_matters
- My pitch is that I can build the product and win its market. Resonate is that pitch pointed at a problem I care about — and the honesty bar is high: it's non-clinical by design, the AI never counsels, and no outcome claims exist until the Fall 2026 cohort produces real data.
- UF Students in Scope
- ~56K
- Core Loop
- < 5 min/day
- Crisis Routing Tiers
- 3
- Launch
- UF · Fall 2026
Resonate turns being a UF student into a daily emotional practice — the non-clinical bridge between everyday campus life and the care students can't get fast enough. You're a Resonator; Echo is your guide; your campus is the game board.
The core loop is deliberately small: check-in (mood + a moment of breath) → a 1–2 minute evidence-based micro-quest → a reflection from Echo → your Resonance Score ticks up. Under five minutes, adaptive, and game-like by design — because the only mental-health product that reaches the students who get no care is one they actually want to open daily. Three deepening loops sit on top: Solo (grounding), Social (VibeMatch — anonymous, resonance-based connection), and Physical (mood 'auras' dropped at real UF buildings; the campus as a living game board).
Safety is built in and non-negotiable: a 3-tier crisis exit runs client-side (instant UI) and server-side (before any AI call), surfacing UF CWC + 988 + 911 — always user-owned, detection supplements and never gates. Resonate is explicitly non-clinical: not therapy, not an EHR, and the AI never counsels. Over time it builds a therapist-ready Journey Map a student can hand to a CWC counselor in one tap — a warm referral pipeline into real care.
Stage, stated honestly: a full-stack v1 was built and deployed solo once, to prove the whole thing is buildable — an execution proof, not a live MVP with users. The v2 prototype is a focused rebuild around that daily loop, code-complete and in launch-gate hardening for a Fall 2026 free learning launch at UF. Everything downstream — a Spring 2027 paid pilot, the institutional intelligence layer, multi-campus — gates on one falsifiable test: voluntary D30 retention materially above the ~3.3% category baseline. No outcome claims exist until that data does.
The business model is institutional B2B: free for students always, licensed per student to the university, sold to Student Affairs. The honest defensibility isn't a network effect or a distribution moat (the university holds the rails) — it's a per-user, compounding switching cost backed by a campus-native data pipeline: eighteen months of one student's mornings and moods isn't reproducible on day one.
- The wedge is a one-sided daily loop that works at n=1 — no two-sided cold-start to survive before there's value.
- Crisis routing has to be layered and user-owned: client-side responds instantly, server-side still guards the AI path, and detection never gates the human exit.
- Say the stage honestly — v1 was an execution proof, not a shipped MVP. Overstating 'shipped' is the fastest way to lose a serious room.
- The whole thesis reduces to one falsifiable question — do students voluntarily come back? — so the launch is a learning launch, not a victory lap.
