HabitPilot

Core Functionality, Prototype Stability & Technical Complexity:
Habit Breaker is a Desktop application that monitors all desktop apps in real time using Electron and active-win. The system connects to a Node.js Express backend server that orchestrates multiple AI services: when users access distracting applications, the backend integrates with Groq LLM (Llama 3.3 70B) to generate personalized intervention messages, and ElevenLabs TTS API creates authentic voice warnings in real time. If users remain distracted, Stripe automatically charges penalty fees from their committed deposits. When daily goals are achieved, UberEats API automatically delivers food as rewards. All data is stored in PostgreSQL. The prototype is stable with a complete multimodal workflow: browser monitoring → cloud-based LLM analysis → voice synthesis → financial incentives → automated rewards → database persistence.
Innovation & Creativity:
The solution combines financial commitment (Stripe), behavioral AI analysis (Groq LLM), real-time voice intervention (ElevenLabs), and tangible rewards (UberEats) into a cohesive productivity system. The innovation lies in the multi-agent orchestration where financial penalties create immediate consequences while rewards provide positive reinforcement, all powered by AI that adapts messages based on cumulative usage patterns.

Real-World Impact:
Addresses digital distraction, which costs billions in lost productivity. The financial incentive model increases accountability, while voice-based interventions are more effective than text-only warnings. Potential applications include corporate productivity programs, student focus management, remote work accountability, and personal digital wellness.
Theme Alignment - Agents from Browsers, Voices, Clouds, and Tools:
Browsers as Agents: Desktop App (Electron + active-win) monitors all desktop applications in real time, detecting which apps users are using and tracking time spent.
Voices as Agents: ElevenLabs TTS API serves as a voice agent that delivers personalized, context-aware interventions with authentic human-like speech, creating emotional connection and higher engagement.
Clouds as Agents: Groq LLM (cloud-based AI) acts as an intelligent agent that analyzes behavior patterns, generates adaptive messages based on cumulative usage, and makes context-aware decisions about intervention severity.
Tools as Agents: Stripe processes financial transactions automatically, UberEats API handles reward delivery, PostgreSQL database stores and manages all user data, and n8n workflow orchestrates automated daily reports—all working together as autonomous agents.
These components form a cohesive multi-agent system: the desktop monitoring agent detects distractions, the cloud AI agent analyzes and generates personalized responses, the voice agent delivers interventions, the financial agent (Stripe) enforces consequences, the reward agent (UberEats) provides incentives, and the database agent maintains persistent state—creating an autonomous productivity assistant that operates without constant human intervention.

Complete Technology Stack:
Desktop Application: Electron, active-win (Node.js module for detecting active windows/applications)
Backend: Node.js, Express.js, REST API
AI & LLM: Groq SDK, Groq API, Llama 3.3 70B Versatile model
Voice AI: ElevenLabs API, ElevenLabs TTS (Text-to-Speech), Real-time voice synthesis
Payment Processing: Stripe API, Stripe Payment Intent, Stripe Charge API
Food Delivery: UberEats API
Database: PostgreSQL
Automation: n8n (Workflow Orchestration)
Email: nodemailer, Gmail SMTP
Storage: electron-store (for Electron app local storage)
Configuration: dotenv
Development: npm, nodemon
Version Control: Git, GitHub