Branch

Inspiration

We wanted to make productivity feel as motivating and measurable as physical fitness. Apps like Strava turn running into a social game of progress, stats, and community — but nothing like that exists for working. We noticed how difficult it is to stay focused, structure long tasks, and feel rewarded for consistent academic effort. Branch was born from the idea that learning deserves a feedback loop — one that tracks focus, gamifies growth, and connects learners socially instead of leaving them isolated behind to-do lists.

What it does

Branch transforms working into a measurable, gamified experience.

Track productivity automatically: A Chrome extension logs activity, classifies sites as productive or distracting using an NLP model, and measures real engagement time.

Generate structured study plans: An LLM breaks large goals (“Finish midterm project”) into smaller, scheduled subtasks on a personalized calendar.

Reward and compete: Users earn points and badges for completing tasks, staying focused, and keeping streaks. They can even bet points on friends’ productivity to make accountability fun.

Social feed: Users share progress posts, send kudos, and generate AI-crafted shareable images of their study milestones.

Leaderboard: See how your focus ranks against your friends or classmates.

How we built it

Frontend: React app for dashboard, leaderboards, and social features.

Chrome Extension: Captures tab activity, interaction signals, and duration; syncs logs to backend.

Backend: Node.js (Express + PostgreSQL) with Redis for leaderboards and real-time updates (Socket.io).

ML Layer:

DistilBERT classifier for activity taxonomy (e.g., coding, note review, video lecture).

XGBoost regression model for computing productivity scores using interaction-weighted time, task weight, and focus index.

OpenAI API for LLM task-splitting and generative share cards.

Infra: Containerized microservices via Docker; Bull jobs for scheduled tasks (badges, reminders, bet resolution).

Data pipeline: Logs → aggregation → scoring → points → badges → leaderboard updates → feed posts.

Challenges we ran into

Designing an accurate productivity metric that goes beyond “time on tab” to reflect real engagement.

Making sure the Chrome extension tracked active windows and idle time reliably without violating user privacy.

Integrating ML components (Python FastAPI service for classification) seamlessly with the Node backend.

Handling real-time updates for leaderboards and bets using Redis and sockets.

Building a consistent user experience that blended study tracking, social motivation, and gamification without overwhelming users.

Accomplishments that we're proud of

Built a fully working prototype connecting the Chrome extension, backend scoring, and React social dashboard.

Created a hybrid ML pipeline (NLP classifier + regression model) that calculates a realistic productivity score.

Implemented a points, badges, and betting economy — a new social mechanic in edtech.

Integrated LLM-based task decomposition to automate study planning.

Designed shareable AI-generated achievement images for social virality.

Achieved near real-time leaderboard updates across multiple users.

What we learned

How to design an ML-driven feedback loop grounded in user data and human psychology.

The importance of privacy and transparency when tracking productivity data.

How to merge gamification with serious academic intent without trivializing the experience.

Building cross-stack: from browser extensions to LLM integrations to real-time backend systems.

The value of beta testing — user feedback drastically improved our scoring model and UI clarity.

What’s next for Branch

Canvas and LMS integrations: Connect with learning platforms like Canvas, Google Classroom, and Notion to automatically import assignments, due dates, and grades for student productivity.

Calendar and timesheet integrations can similarly provide further data for business related productivity.

Expanded productivity metrics: Use beta testing feedback to refine the ML model with richer behavioral signals (typing cadence, focus interruptions, context switching).

Mobile companion app: Bring live productivity tracking, reminders, and social feed to mobile.

Institutional pilots: Partner with universities and study communities to deploy Branch as a study motivation tool.

Enhanced community gamification: Work challenges, group leaderboards, and verified study clubs.