Track: AI for Learning & Developer Productivity
Built for: AI for Bharat Hackathon 2026
📖 Overview The Deep-Work Potential Predictor is an AI-powered productivity ecosystem designed to help students and developers master their focus. By predicting your Focus Score (0-100) using biometric and environmental data, the app intelligently schedules your hardest tasks (like debugging or math) during your peak cognitive windows.
🗺️ User Journey & UI/UX Progress "The user starts by inputting biometric data (sleep, nutrition) on a clean dashboard. Our AWS Lambda backend processes this to display a real-time 'Focus Score' powered by SageMaker. Based on this, Amazon Bedrock generates a personalized task roadmap, and users earn 'Deep-Work Points' (DWP) to level up their rank from Novice to Legend." See the full wireframes in the /Design folder.
Key Features Predictive Scheduling: Uses Multiple Linear Regression to calculate your "Flow State" probability.
AI Roadmap Generator: Converts long-term goals (e.g., Andrew Ng’s ML course) into manageable daily "bits."
Smart To-Do List: Automatically prioritizes tasks based on your predicted focus score.
Focus Guard: Integrated app-blocking logic to minimize distractions during high-focus periods.
🛠 Tech Stack This project is built following the AWS Well-Architected Framework:
Generative AI: Amazon Bedrock (Claude 3.5 Sonnet) for Roadmap Synthesis.
Machine Learning: Amazon SageMaker for Focus Score regression modeling.
Compute: AWS Lambda (Serverless Backend).
Database: Amazon DynamoDB for persistent storage of user vitals.
Dev Tools: Amazon Q & Kiro AI IDE (Mandatory Spec-Driven Development).
📐 Spec-Driven Development To ensure architectural integrity, this project strictly followed the Kiro workflow:
requirements.md: Functional specs defined using EARS notation.
design.md: Technical blueprint and system data flow.
.kiro/: Mandatory compliance directory tracking the development lifecycle.
🚀 Getting Started Clone the repo:
Bash git clone https://github.com/IshaanMogs/Deep-Work-Potential-Predictor.git Review the Specs: Check requirements.md and design.md to understand the system logic.
Authors Ishaan Kumar - Lead Developer & Architect