Personal Twin - Team 26

🚀 About the Project: AI Digital Twin Creator 🌟 Inspiration

This project was inspired by the idea that a person’s professional identity is more than a résumé—it’s a living, evolving story. I wanted to build a system that doesn’t just store your experience but can speak, think, and interact like you. The concept of a “digital twin” felt like the perfect bridge between personal data, AI analysis, and real-time communication.

🧠 What I Learned

Throughout the project, I gained a deeper understanding of:

Integrating multiple AI platforms (Fastino, OpenRouter, ElevenLabs) into a single workflow

Real-time data pipelines and background process scheduling

Frontend–backend communication using React, TypeScript, and Python

Efficient caching, API usage optimization, and persistent state management

Designing modular architectures that allow the AI twin to evolve automatically

This helped evaluate when to re-analyze content and when cached summaries were sufficient.

🛠️ How I Built It

The project is composed of three major components:

Backend (Python)

Built a GitHub scraper that analyzes repositories and generates summaries using GPT-based models

Developed an upload and synchronization pipeline for both Fastino and ElevenLabs knowledge bases

Implemented a background scheduler that checks for new repositories every minute

Added smart caching and consolidated JSON outputs for fast retrieval

Frontend (React + TypeScript)

Designed a modern UI for uploading resumes, connecting GitHub, and chatting with the AI twin

Integrated a real-time chat interface powered by Fastino

Embedded ElevenLabs’ voice widget for conversational AI in natural speech

Built smooth, animated interactions to make the experience feel polished and intuitive

Automation Layer

Created a persistent system that tracks previously analyzed repositories

Automatically updates both knowledge bases when new content is detected

Ensures no duplicate uploads and merges past and new documents seamlessly

⚠️ Challenges I Faced

Building this project came with several obstacles:

Managing Multiple AI APIs

Each service had its own rate limits, authentication system, and response formats. Harmonizing them into a single cohesive pipeline required careful design.

Real-Time Monitoring

Running a background job every minute without blocking the main API took optimization and threading/scheduling considerations.

Data Consistency Across Platforms

Ensuring that Fastino, ElevenLabs, and local storage all stayed synchronized—especially when new repositories were detected—was tricky. I had to build custom logic to avoid duplicates and maintain a clean knowledge base.