STARTED

STARTED - Icon

Inspiration

We picked Gabby's brief because action paralysis hits home for us. That feeling of aiming high, getting excited about a goal, then getting stuck in your own head — overthinking instead of doing. Our team is a mix of men and women, but we think this struggle is universal. We wanted to build something that actually helps.

Before writing code, we talked to women in the target demographic (ambitious, 25-35) to validate the idea. One insight kept coming up: they didn't want another app with a long TODO list and a rigid plan that just adds to the overwhelm. That made us commit to the core design decision — show one action at a time.

We also went deep on the research. We listened to podcasts, read books like Atomic Habits, and studied frameworks for goal achievement. One approach that resonated was Reverse Goal Setting — start with the dream, define how it should make you feel, identify the skills and habits you need to become the person who can achieve it, surface the obstacles in your way, and then focus on just 2-3 actions at a time before reviewing and adjusting. We combined this with the four laws of behavior change from Atomic Habits to build the framework behind the app. The AI uses all of this — but we made it simpler for the user: one action, one day at a time.

On a personal note, I (Eric) am living this right now. I quit my day job to build something of my own. I know I can make it happen, but it's overwhelming. You battle impostor syndrome daily. You're constantly doing things you've never done before — which grows your skills, but you're never quite sure how close you are to making the dream real. STARTED is an app that I think could help me out in this new journey.

## What it does

STARTED gives ambitious women one small, personalized action each day — what we call a "Power Move" — to close the gap between big dreams and real progress.

You set your North Star (your big goal), and Abby, your AI companion, breaks it down into daily micro-actions tailored to you. Complete your move, collect stars, and watch your progress visualized as a constellation you're building toward your dream.

No overwhelm. Just show up, make your move, and level up.

## How we built it

Mobile app: React Native + Expo. We started with Rork to quickly prototype the onboarding flow, then shifted to vibe-coding with Claude Code for the rest.
Backend: Bun + Fastify + Prisma with PostgreSQL.
AI: OpenRouter as the provider, Claude Sonnet 4 powering Abby's personality and the Power Move generation engine.
Monetization: RevenueCat for subscriptions. We went with a hard paywall after a 3-day free trial — for a new app with AI costs, we believe you should focus on a single clear offer and iterate. Pricing is $1.99/week or $49.99/year (~50% discount). At our estimated AI cost of ~$5/year per daily active user, even the annual plan gives us healthy margins to sustain and improve the product.
Assets: AI tools for generating sprites and video assets.
Timeline: Research started around Jan 23. First line of code was Feb 3.

## Challenges we ran into

The biggest challenge was the timeline. We wanted a quality app we'd actually be proud of — not just a hackathon demo.

We had to stay flexible. When something got too complicated, we'd step back and ask: "What other solution could bring 80% of the outcome with 20% of the effort?" That discipline kept us shipping.

RevenueCat itself wasn't painful (prior experience helped), but Apple's subscription review process is... Apple's subscription review process.

## Accomplishments that we're proud of

The Power Move loop engine. The AI system that generates personalized daily actions is the heart of the app. It's a first implementation, but we think it has real potential to get even smarter.
The overall UX. We're genuinely surprised how good the app looks and feels. The dreamy star-filled aesthetic, the animations, the flow — it came together better than expected. We need real users to validate how well it works in practice, but we're proud of the experience we crafted.
The progress visualization. Showing your completed moves as a constellation path toward your North Star was a tough design problem. We're happy with where it landed, and it's one of the features we're most excited to keep iterating on.

## What we learned

Get user feedback faster. We did one round of validation before building, but ideally we would have done more iterations on the core feature with real users. Tough to do in a 4-week hackathon, but it's the lesson we're taking forward.
New tools: First time using some AI tools for asset generation — sprites, video. Opened our eyes to how much faster you can move now.
Scope discipline: The 80/20 question saved us multiple times. Ship the simpler version, learn, then improve.

## What's next for STARTED

If we partner with Gabby, we'd collaborate on a marketing strategy with her audience and incorporate her feedback into the app.

On the product side:

User control over the plan — Right now Abby generates your Power Moves. But users should be able to steer too — suggest what they want to work on, add specific milestones. The AI should adapt, not dictate.
Life context & constraints — What if you have kids and your available time just shifted? We want users to tell Abby about their reality so the plan adjusts to them — not the other way around.
Retention & engagement — Push notifications, home screen widgets, and social sharing with branded share cards for milestones and goal completions.
Foundation — User authentication for cross-device sync, analytics, and a proper feedback system.