Guidr

Guidr transforms AI into structured coaching systems. Access immediate, affordable guidance through beautiful, step-by-step "recipes" for productivity and growth.

Comment

The Guidr Story: Building Systemized AI Coaching

The Inspiration: From Overwhelm to Guided Action

We were immediately drawn to Simon's brief because it addressed a profound gap we've all felt: the struggle to translate inspiration into systematic action. His audience—productivity enthusiasts who value systems—wasn't being served by generic AI chatbots. They needed structure, not just conversation.

Simon’s concept of "coaching systems" resonated deeply. We realized that the true innovation wasn't in creating another AI, but in productizing the methodology of coaching itself. This led to our core concept: Guidr (pronounced "guide-er")—not an AI chatbot, but a platform for systemized coaching "recipes."

How We Built It: A Full-Stack Architecture in 4 Weeks

The Technical Stack

We engineered a scalable, full-stack mobile application using modern tools chosen for speed and reliability:

· Frontend: React with TypeScript, wrapped with Capacitor for native iOS/Android deployment · Backend: Firebase Ecosystem (Authentication, Firestore, Cloud Functions) · AI/ML: Gemini 1.5 Flash API with carefully engineered prompts · Monetization: RevenueCat SDK for cross-platform subscription management · Design: Custom CSS with a minimalist aesthetic matching Simon's brand

The "Coaching Recipe" Engine

Our technical breakthrough was abstracting coaching methodologies into structured prompt templates. Each Guidr (like "The Weekly Review System") is essentially a JSON-defined conversation flow that the AI follows religiously.

// Example of our prompt engineering structure
const weeklyReviewPrompt = {
  systemRole: "You are a focused, systems-oriented productivity coach...",
  steps: [
    "Celebration: What's one win from your past week?",
    "Friction: What was one recurring challenge?",
    // ... structured dialogue flow
  ],
  rules: "Ask ONE question at a time. Do not deviate from steps."
};

Monetization Architecture

We implemented a freemium model with clear value progression:

· Free tier: One complete Guidr (Weekly Review) · Premium tier ($4.99/month): All Guidrs + advanced features

The business case was straightforward: if a traditional coach costs $200+/hour, Guidr provides ongoing value at approximately 1% of the cost:

\text{Value Ratio} = \frac{\text{Guidr Premium}}{\text{Traditional Coaching}} = \frac{\$4.99}{\$200} \approx 2.5\%

Challenges We Overcame

  1. The AI "Over-Creativity" Problem

Early versions would skip steps or offer unsolicited advice. Our solution was strict prompt engineering with explicit constraints and state tracking through conversation history.

  1. Cross-Platform Payment Complexity

Implementing subscriptions that work seamlessly on both iOS and Android required deep integration with RevenueCat's SDK and careful testing of purchase flows.

  1. Data Structure Design

Balancing Firestore's NoSQL flexibility with the need for structured conversation history led us to implement a hybrid approach—structured prompts with flexible conversation storage.

What We Learned

Technical Insights

· Prompt engineering is a distinct skill from general AI development · Firebase Cloud Functions provide an excellent serverless backbone for AI applications · RevenueCat dramatically simplifies what would otherwise be months of payment infrastructure work

Product Insights

· Constraints drive creativity: Limiting the AI to specific frameworks made it more useful, not less · Personalization matters: Even simple context (user's goals, weekly sentiment) dramatically improves perceived value · Mobile-first changes everything: Push notifications and native performance create stickiness web apps can't match

The Future of Guidr

We've built more than a hackathon submission—we've created a platform for coaching democratization. The architecture allows any expert to create and share their methodologies as Guidrs, potentially creating a marketplace of vetted coaching systems.

The technical foundation is ready for:

· Community Guidr creation tools · Progress analytics and insights · Integration with productivity tools (Calendar, Todoist, Notion) · Group coaching sessions

Conclusion

Guidr represents what's possible when you combine AI capability with human wisdom, structure creativity with systems, and build technology that serves rather than overwhelms. We didn't just build another AI app—we built a bridge between the promise of AI and the real human need for guidance.

For Simon's audience of system-builders and productivity enthusiasts, we believe Guidr offers exactly what they crave: AI that works within their systems, not outside of them.

Built with React, Firebase, Gemini, and RevenueCat for the 2026 RevenueCat Shipyard Hackathon.

Built With

Share this project:

Updates