Back to Projects

Focus — Behavioral AI Productivity Platform

4 microservices, 9-gate AI insight pipeline, 5 automated CI/CD workflows

A production-grade, AI-powered productivity platform that learns user behavior and delivers personalized coaching. Built with a 4-service Spring Boot microservices architecture, API Gateway, and a Flutter mobile client. It combines focus sessions, smart task tracking, notes, user streaks/XP, screen-time analysis, and AI-generated insights that adapt to actual habits.

Flutter 3.xDart 3.10Java 17Spring Boot 3.5.xSpring Cloud GatewayMongoDB AtlasGemini APIJWTDockerKubernetes (OpenShift)GitHub Actions

Project Screenshots

Focus — Behavioral AI Productivity Platform screenshot 1

Problem

Most productivity apps are passive timers that count down. Users need an active behavioral coach that understands their actual usage patterns, goals, and bad habits to provide contextual, personalized nudges rather than generic advice.

Solution

Developed a full-stack system with a Flutter app and a Spring Boot microservices backend. Integrated a 9-gate insight pipeline featuring a Rule Engine and Gemini AI with dynamic fallback. Features native Android screen-time syncing, JWT rotation, BCrypt OTP, account lockout, and rate limiting.

Key Highlights

  • 3-layer behavioral AI (Micro, Expanded, Confidence) using Gemini API with dynamic fallback
  • 4 microservices + API gateway architecture for scalability and separation of concerns
  • Enterprise-grade security: JWT rotation, BCrypt OTP, account lockout, rate limiting
  • Native Android bridge for screen-time collection fed directly into AI prompts
  • Gamification system with XP, levels, and streak tracking
  • Automated CI/CD pipeline deploying to Kubernetes (Red Hat OpenShift) via GHCR