Aura: Minimalist AI Coaching

Aura: Minimalist AI Coaching

Comment

Aura: The Minimalist AI Coaching Ecosystem

⛴️ Shipyard: Creator Contest — Submission Report

Project Overview

Aura is a high-fidelity mobile MVP designed specifically for Simon (Better Creating) and his audience of intentional, productivity-focused creators. Aura transforms the often-overwhelming landscape of General AI into a sanctuary of focused, minimalist guidance. It isn't just a chatbot; it's a context-aware framework for personal growth that bridges the gap between raw information and actionable wisdom.

Video Demo

Watch the Aura Walkthrough Here Note: This is a placeholder link for the required 2-3 minute walkthrough.

1. The Inspiration

The inspiration for Aura stems from a fundamental problem in the modern "Productivity Porn" era: Information Overload.

Simon’s audience loves building systems (Notion, Obsidian, GTD), but they often struggle with the "friction of start." General-purpose AI tools like ChatGPT are too broad; they lack the specific context of a user's deeply held values and current life goals.

I wanted to build an app that felt like a physical notebook: quiet, intentional, and high-quality. The design language takes cues from "Calm Technology," ensuring the interface recedes to let the user's thoughts and the coach's guidance take center stage.

2. How It Was Built

Technical Architecture

Aura is built using a modern, reactive stack designed for speed and modularity:

  • Core Framework: React 19 with TypeScript for type-safe state management.
  • Styling: Tailwind CSS for a utility-first, minimalist design system.
  • AI Engine: Google Gemini API (gemini-3-flash-preview) utilizing advanced system instructions.
  • Monetization Architecture: A simulated RevenueCat integration layer that manages entitlements and paywall states.

Context-Aware Intelligence

The core "magic" of Aura is the Context Injection Engine. Every message sent to the AI is wrapped in a dynamic system prompt:

$$P_{final} = S_{coach} + C_{user} + M_{history} + U_{input}$$

Where:

  • $S_{coach}$: The specific persona and specialty of the selected mentor.
  • $C_{user}$: The user's specific values and goals stored in the "Context" settings.
  • $M_{history}$: The short-term conversational memory.

Monetization Strategy (RevenueCat Integration)

Following the RevenueCat "Shipyard" requirements, Aura features a multi-tiered subscription model:

  1. Free Tier: Access to basic coaches (e.g., The Productivity Sensei).
  2. Aura Pro ($49.99/year): Access to "Specialist" coaches like the Financial Pathfinder or Career Architect.

The implementation simulates the RevenueCat SDK's offerings and purchaserInfo logic, ensuring that premium content is gated behind a high-converting, beautiful paywall that supports "Restore Purchases" functionality.

3. Challenges Faced

The Minimalist Paradox

The biggest challenge was the Minimalist Paradox: How do you make an app feel "feature-rich" while removing almost all UI elements?

  • Solution: We used motion and micro-interactions. Instead of heavy menus, we used subtle haptic-like animations and transitions. If a user clicks a coach, the card "breathes" before navigating, providing feedback without clutter.

Prompt Drift

Initial tests showed the AI sometimes forgot the user's specific goals mid-conversation.

  • Solution: I implemented a "Refresher Block" in the system instruction that forces the model to evaluate the user's currentGoals array before every response.

Monetization Friction

Paywalls are traditionally annoying.

  • Solution: I designed the Aura paywall to feel like an extension of the app's aesthetic—using dark modes, high-quality gradients, and clear typography to reduce "buyer's remorse" and increase trust.

4. What I Learned

  1. Prompt Engineering as UI: I learned that in AI-first apps, the "backend" prompt is just as much a part of the UX as the "frontend" button. A poorly phrased system instruction can ruin a beautiful design.
  2. The Power of Constraints: Building for a specific influencer (Simon) helped narrow the scope. Having a "target audience" (minimalist creators) made every design decision easier because I could ask: "Would Simon find this too noisy?"
  3. Monetization is UX: Integrating RevenueCat patterns taught me that the purchase flow is the most sensitive part of the user journey. It must be transparent, fast, and respectful.

5. Mathematical Framework of Coaching Value

To quantify the effectiveness of the Aura system, we look at the Coaching Value Formula:

$$V = \frac{C \cdot A}{F}$$

Where:

  • $V$ is the perceived value to the user.
  • $C$ is the Context Accuracy (how well the AI knows your goals).
  • $A$ is the Actionability of the response.
  • $F$ is the Friction (UI clutter, slow response times, complex setup).

By maximizing $C$ and $A$ while minimizing $F$ to near-zero through minimalist design, Aura achieves a higher $V$ than standard chat interfaces.

6. Future Roadmap

If selected for the next phase, the following features are planned:

  • Audio Mode: Real-time voice coaching using Gemini Live API.
  • Context Auto-Sync: Integration with Notion/Obsidian to automatically update goals.
  • RevenueCat Experiments: Implementing A/B testing on paywall designs to optimize conversion rates for the influencer.

7. Conclusion

Aura is more than an entry into a contest; it is a prototype for the future of intentional AI interaction. By combining Simon's philosophy of "Better Creating" with the power of the Gemini API and the reliability of RevenueCat, we have created an MVP that is ready to ship, monetize, and scale.

Anchors aweigh! ⚓️

Built With

Share this project:

Updates