Clairity Daily

Product Strategy: The "Business Thinking"

I didn't leave the corporate world to settle for another 9-to-5. After my time at Corteva, I realized I was craving more than just a paycheck—I wanted to work with passionate builders who are obsessed with solving real problems. I saw a massive disconnect in how people use AI: most were just "prompting," while a select few were "thinking" natively in AI.

I set out to close that gap. My mission wasn't just to build an app; it was to empower people to live happier, healthier, more fulfilling lives by mastering the tools of the future.

To do that, I had to walk the walk. I spent 70 days in a self-imposed "AI Bootcamp," testing one question: Do I have what it takes to build entire full-stack apps on my own? I wasn't just building a product; I was designing a career path defined by action, velocity, and impact.

The Problem

I saw two barriers holding people back:

Pitching Clairity Daily

The "User" Trap: Most people are passive consumers of AI. They ask a question and accept the answer. I wanted to turn them into active engineers of their own tools, teaching them to reverse-engineer the "black box" of LLMs.
The "Founder" Trap: Too many people (myself included) wait for permission or the "perfect team" to build their ideas. I needed to prove that with the right focus, you can bridge the gap between "idea" and "execution" entirely on your own.

The Build

This was pure velocity. I compressed a year's worth of learning into a single summer.

The Game: I created "Bridge the Gap," a daily challenge that gamified AI literacy. Instead of consuming content, users had to deconstruct it—guessing the prompts and models behind the output. Each day, an admin-seeded scenario and AI-generated output were published. Players selected a model (GPT-4o-mini, Llama 3.1, or Gemini 2.0 Flash) and wrote a prompt to recreate it. Scores were computed in real time using cosine similarity between embeddings of the player's output and the goal output — weighted across model selection (10 pts), prompt similarity (50 pts), and output similarity (40 pts).
The AI Dev Suite: I tried every AI-accelerated tool I could find—Cursor, Windsurf, Claude Code, and v0—to write production-grade code at lightning speed while maintaining security and architecture control.

The Bridge the Gap game interface I designed

Full-Stack Execution: I didn't just write the Next.js code. I designed the artwork, built the Supabase backend, set up the Vercel AI Gateway, and even hit the podcast circuit to promote it. I was in the arena every single day, turning a vision into a tangible reality.

How It Was Built

The stack was chosen for speed and production credibility — nothing experimental, everything integrated. See the full breakdown in ARCHITECTURE.md.

Example schema for Bridge the Gap

The architecture separated game data into five Supabase tables: public game metadata, a protected goal output, the goal prompt and model, pre-computed similarity vectors, and per-user submissions and scores. Clerk JWTs were wired directly into Supabase as the access token, with a separate admin client for privileged server-side operations.

Access was gated by date: today's game was free for any signed-in user, past games required a paid subscription enforced via Clerk's play_past_games feature flag.

The Outcome

The app works perfectly, but the real product was me.

Lifetime Users - Clerk

The Reality Check: While the code was clean and the design was slick, users didn't stick around. I learned the most valuable lesson of startups: Validate before you build. Passion is the fuel, but the market is the steering wheel.
The Pivot to Partnership: This "successful failure" got me into Ignition, a competitive startup incubator at 16 Tech. There, I realized my true strength isn't just coming up with ideas—it's executing them.

Presenting at Ignition startup incubator at 16 Tech

A New Mission: I now use my technical velocity to help other founders build their dreams. I've realized that I'm happiest when I'm the technical partner who brings the visionary idea to life.
The "Why": Clairity Daily didn't become a unicorn, but it led me to a life I love. I now spend my days helping others create that sense of joy and wonder, ensuring they build scalable, value-driven products that actually make a difference.

What I Learned

Infrastructure should follow signal, not precede it. I built a stateful user system, persistent scoring, RLS policies, and an embedding pipeline before validating whether people wanted to come back the next day. A stateless prototype with no auth and no database could have tested the core game loop in a fraction of the time.

Separate "can I build this?" from "should this exist?" This project answered the first question definitively — yes, I can design, build, and deploy a full-stack AI product solo. But the second question requires earlier and more focused validation. I invested in infrastructure, UX polish, and system design before confirming daily retention or willingness to return.

A deliberate shutdown is a good outcome. Continuing would have meant deeper product validation, ongoing infrastructure maintenance, and a longer-term commitment without a clear signal. The highest-ROI decision was to archive the project, extract the lessons, and move forward. That's not failure — that's judgment.