Mentara

Your personal "Board of Directors": A minimalist AI coaching system that replaces generic advice with specialized, methodology-driven micro-coaches.

Comment

Inspiration

We are huge fans of Simon (@BetterCreating) and his philosophy that "You don't rise to the level of your goals, you fall to the level of your systems."

However, we noticed a critical gap in the current AI landscape. Most AI assistants are "generic generalists." They try to be everything to everyone. When you ask a generic LLM for advice, you often get a bland, average answer.

Simon’s brief asked for "Minimalist AI coaching in your pocket" where users can "browse, create, and share AI coaches". This sparked our core realization: High-performers don't need one coach; they need a Board of Directors. Sometimes you need a ruthless prioritizer; other times, you need an empathetic listener. Mentara was born to replace the "blank cursor anxiety" with immediate, methodology-specific guidance.

What it does

Mentara is a minimalist "Micro-Coach Operating System." Instead of chatting with a single bot, users select from a deck of specialized AI Personas, each trained on a specific productivity framework.

  • The Deck: Users swipe through beautifully designed cards to choose their "mode" (e.g., The Deep Worker, The Stoic, The Gregarious Motivator).
  • Context-Aware Chat: We implemented a "System Context" engine. The user inputs their core values and current goals once, and every coach automatically knows this context—eliminating the need to re-explain yourself.
  • Frictionless Monetization: We used RevenueCat to build a "Freemium" model. Users get access to the "Generalist" coach for free, but must subscribe to unlock the specialized "Pro" coaches (like the Executive Function Specialist).

How we built it

We prioritized speed and design fidelity to meet the "clean, minimal" requirement of the brief.

  • Frontend: We built the mobile app using Flutter to ensure a high-performance, native feel on both iOS and Android. The UI is strictly minimalist—no clutter, just focus.
  • AI Engine: We utilized the OpenAI API, designing complex "System Prompts" for each coach. We engineered specific "temperature" and "tone" settings so "The Ruthless Prioritizer" actually refuses to engage in small talk, while "The Empathetic Reflector" uses active listening techniques.
  • Monetization: We integrated the RevenueCat SDK to handle subscriptions. This allowed us to set up a Paywall in under an hour, letting us focus on the product experience rather than backend billing infrastructure.

Challenges we ran into

  1. Balancing "Minimalism" with "Power": Simon's brief emphasized "No complex setup". We initially wanted complex forms for user onboarding, but realized that killed the magic. We stripped it down to just three fields: Name, Core Value, and Main Goal.
  2. Prompt Engineering nuances: It was difficult to stop the AI from breaking character. "The Ruthless Prioritizer" kept trying to be polite! We had to use strict system instructions to force it to be concise and direct, ensuring each coach felt like a distinct personality, not just ChatGPT with a different name.

Accomplishments that we're proud of

  • We are incredibly proud of how distinct the coaches feel. Switching from one coach to another changes the entire vibe of the advice, which feels magical.
  • We went from "zero" to a working "Pro Subscription" flow in record time.
  • In a 4-week hackathon, UI often gets neglected. We are proud that Mentara looks and feels like a premium product that belongs on the App Store today.

What we learned

  • Context is King: AI is 10x more useful when it knows who you are before you speak.
  • Niche > General: A tool that does one thing (coaching via specific methods) is far more valuable than a tool that does everything.
  • Monetization is a Feature: Building the paywall with RevenueCat forced us to think about value. We had to ask ourselves, "Is this specific coach valuable enough to pay for?" It sharpened our product thinking.

What's next for Mentara

  1. Creator Economy: We want to allow creators (like Simon!) to build their own AI Coach personas (trained on their specific YouTube videos and notions templates) and sell them on the Mentara platform.
  2. Voice Mode: Adding real-time voice interaction for on-the-go coaching.
  3. Wearable Integration: Pushing "Micro-Interventions" to smartwatches to keep users on track with their systems throughout the day.

Built With

Share this project:

Updates