Catalog of work · 2023 — 2026
Three products,
shaped from zero.
Every product here began as a sentence and a feeling. The work below is what it took to make them buildable.
Consumer AI · Wellness · 2023 — Present
Harmoni
An AI-powered wellness platform.
● Live
- Founder
- Product strategy
- PRD authorship
- Roadmap
Harmoni · Plate 01
Harmoni.
An AI-powered wellness platform.
Harmoni began as a personal frustration: wellness apps that confused content for change. I scoped a platform that uses AI to guide real progress — personalized, evidenced, and quietly intelligent.
The thesis
Wellness software fails when it optimizes for streaks instead of outcomes. The thesis: pair adaptive AI guidance with structured personal practice, and ship the smallest version that proves a single user behavior change.
What I scoped
A documented product vision, three user personas, a 14-feature MVP scope (reduced from 41), monetization tiers, and a PRD the engineering team could quote against. We shipped the MVP in 11 weeks.
Outcome
Harmoni is live with paying users and continues to evolve. Every feature shipped traces back to a written user story — no orphan code, no unexplained surfaces.
- Status
- Live in production
- Scope
- 14 MVP features
- Cycle
- 11 weeks to ship
- Stack
- Web + iOS + AI gateway
B2B Healthtech · 2024 — Present
Harmoni Care
A clinician portal for wellness professionals.
● Integration & Testing
- Product lead
- PRD
- Stakeholder discovery
Harmoni Care · Plate 02
Harmoni.
A clinician portal for wellness professionals.
Care extends Harmoni into the clinician's workflow — letting therapists and wellness professionals support clients with the same intelligence layer their clients already use at home.
The thesis
Practitioners want continuity, not another dashboard. The product had to mirror Harmoni's consumer language while introducing the smallest possible learning curve for clinical workflows.
What I scoped
Personas for three clinician types, a permissions model that respects client privacy, an MVP focused on session prep and between-session insight, and a PRD reviewed by two practicing therapists before a single ticket was opened.
Outcome
Currently in integration and testing. A small cohort of practitioners is shaping the production release.
- Status
- Integration & testing
- Audience
- Licensed practitioners
- MVP
- Session prep · insight feed
- Phase
- Closed practitioner cohort
AI · Atmospheric · 2025
Meso
AI for weather intelligence and storm chasing.
● Beta Testing
- Product strategy
- MVP definition
- Founder support
Meso · Plate 03
Meso.
AI for weather intelligence and storm chasing.
Meso is built for the people who chase weather rather than check it. The product translates atmospheric data into actionable forecasts for operators in the field.
The thesis
Generic weather apps optimize for the commuter. Meso optimizes for the operator: storm chasers, field photographers, atmospheric researchers, and emergency teams who need confidence, not vibes.
What I scoped
Defined the operator persona, narrowed the MVP to a single high-stakes use case (storm cell intercept guidance), and built the feature priority matrix that protected the founder from feature creep during AI integration.
Outcome
In active beta with field operators. Iterating against real-world conditions, not synthetic ones.
- Status
- Closed beta
- Use case
- Storm cell intercept
- Cohort
- Field operators
- Next
- Public beta · 2026