Sage - AI Coach

AI coaching that actually knows you. Sage remembers your context, adapts to your journey, and delivers personalized guidance that gets smarter over time

Comment

## Inspiration

Like many people who rely on AI daily, I kept running into the same frustrating problem: every new conversation starts from zero. Switch models, switch apps, or even start a new chat — and suddenly the AI has no idea who you are, what you're working on, or what you've already discussed. I found myself copy-pasting context over and over, especially as someone who switches between multiple AI providers. That personal pain point became the foundation for Sage — an AI coaching app that actually remembers you. When I came across Simon's BetterCreating brief and his vision for accessible, calm productivity tools, the fit felt natural. His philosophy aligned perfectly with what I wanted to build: something simple, smart, and genuinely personal.

## What it does

Sage is a personal AI coaching app with persistent memory. It learns who you are — your goals, preferences, communication style, and context — and carries that knowledge across every conversation. Unlike traditional AI assistants that forget you after each session, Sage ensures every interaction picks up right where you left off.

Users get access to specialized AI coaches for productivity, decision-making, goal-setting, content writing, and more — available anytime, with no appointments or scheduling. No waiting for a coach's availability like in real-life coaching. And with multi-model support (OpenAI, Claude, Grok), users aren't locked into a single AI provider.

Beyond personal use, Sage features a community marketplace where users can share their best-performing coaches with friends or discover top-rated coaches created by others — turning great coaching setups into something everyone can benefit from.

## How we built it

The core idea was clear from day one: build a chat system that truly understands you. But as I used the app throughout development, new needs emerged organically. That's how features like coach sharing and the community marketplace came to life — born from real usage, not just planning.

For the tech stack, I chose Flutter and Firebase — a combination I've mastered through years of side projects, which let me move fast without sacrificing quality. RevenueCat has been my go-to for subscriptions across all my projects, so integration was seamless.

The design was inspired by Google Stitch, aiming for a balance between minimalist productivity aesthetics and a modern, smart feel fitting for a coaching product. Warm gradients, glassmorphism, and clean typography create an experience that feels calm yet sophisticated — aligned with Simon's "ambient calm" design philosophy.

## Challenges we ran into

The biggest challenge was scope management. With a short contest timeline, ideas kept flowing faster than I could build them. Deciding what to include and what to save for later required real discipline.

Video production was another hurdle. As a developer with zero experience in making demo videos, learning to communicate the product's value visually — not just technically — pushed me well outside my comfort zone.

## Accomplishments that we're proud of

The user experience feels cohesive in a way that's rare for a contest-timeline project. Everything is connected — every coach exists for a purpose, navigation feels intuitive, and the design is simple yet polished. It doesn't feel like a hackathon prototype; it feels like a product people would actually want to use.

I'm also proud of how well the app aligns with Simon's guidelines and design philosophy. It's not just an app built for his brief — it genuinely embodies his vision of accessible, calm, and personal productivity tools.

## What we learned

Vibe coding is a new era. This app would have taken months to build using traditional workflows. With AI-assisted development, it was fully production-ready in just three weeks — from concept to TestFlight.

I also discovered RevenueCat's Experiments feature, which I hadn't used before despite being a long-time RevenueCat user. What stood out is the flexibility in setting experiment goals — whether it's conversion rate, total revenue, or user LTV — which is more targeted than the A/B testing tools I was using previously.

## What's next for Sage - AI Coach

I hope Simon will love the app. A partnership with him would be a dream — his audience is exactly who Sage is built for, and his endorsement could take the app to the next level. Beyond the contest, the roadmap includes expanding the coach marketplace with community-driven contributions, implementing conversation insights so users can track their growth over time, and continuing to refine the personalization engine that makes Sage unlike any other AI tool out there.

Built With

Share this project:

Updates