Overview
Turning a fragmented learning product into one repeatable mastery ritual
I redesigned and shipped Mindfocus around one clear loop: Memento → FlashTest → mastery.
7 → 2
steps to first value
Context
A memory product whose core value was hidden
Mindfocus helps users turn a subject into knowledge they can recall without assistance. The original product offered more than ten tools, but the path from arrival to mastery was difficult to understand.
Starting point
More than ten features competed for attention on the home screen
Seven steps separated a new user from the first value moment
The Memento → FlashTest → mastery loop was buried in the interface
Mechanic
One repeatable ritual instead of a collection of features
The product was rebuilt around one sequence: capture what matters, structure it, test recall, make mastery visible, then return when memory needs reinforcement.
Decisions
Make the product promise immediately readable
The fragmented homepage became a focused entry point into one learning ritual. Secondary tools stopped competing with the action that creates value.
Fragmented learning experience — before
Single learning ritual — after
Reduce the path to first value from seven steps to two
The onboarding wizard was removed. Creating a Memento became the first interaction, so users could experience the product before learning its entire system.
Multiple onboarding steps — before
Single step to first value — after
Turn mastery into visible product feedback
FlashTest moved from a standalone exercise to the proof mechanism for the whole experience. Results now expose progress and guide what the user should reinforce next.
Standalone FlashTest — before
Mastery validation system — after
System
A product mechanic backed by a working production system
I designed and implemented the systems behind the experience—from memory scheduling and AI-assisted mastery to real-time collaboration and immersive 3D learning journeys.
Memory engine
MongoDB persists each Memento’s mastery state, review schedule and timezone. Scheduled jobs drive the reinforcement cycle from J0 to J365.
AI-assisted mastery
A seven-step generation pipeline creates each Memento. Audio mastery sessions are transcribed and evaluated server-side before the memory state can advance.
Real-time rooms
Socket.IO synchronizes rooms, participant presence and timers. WebRTC powers peer-to-peer audio, video and screen sharing.
Immersive 3D journeys
Three.js turns collections and mastery progress into spatial learning environments, with custom transitions and interaction logic across the journey.
Production foundation
NextAuth handles magic-link and Google authentication, Stripe manages subscriptions, PostHog measures product usage, and AWS S3 stores user media.
Next.js · React · MongoDB · OpenAI · Socket.IO · WebRTC · Three.js · NextAuth · Stripe · PostHog · AWS S3
Impact
One loop now connects first use, mastery and return
The redesign aligned positioning, interaction and implementation around the same promise: create a Memento, prove mastery, and know when to return.
The product promise is readable from the first screen
Memento creation became the first meaningful action
Mastery became visible, measurable and scheduled over time
Core activation and return behaviour are tracked through product analytics
