Project case study
TruthLayer
Analyze online content and surface supporting context with an AI-assisted misinformation detection tool.
Top 6 hackathon project
Project status
Completed
My role
Full-Stack & AI Developer
Timeline / Year
2025
Project Overview
TruthLayer is an AI-assisted misinformation detection platform built during a hackathon to help users evaluate questionable online content. Through a browser extension, users can analyze posts while browsing and receive credibility estimates supported by retrieved reference material and contextual information.
Key Features
- AI-assisted credibility analysis for online content
- Chrome extension for analyzing Faceb posts while browsing
- Machine learning classification for misinformation detection
- Retrieval-augmented verification using vector search
- Supporting context and references alongside credibility estimates
Technical Overview
Tech stack
- TypeScript
- Python
- FastAPI
- scikit-learn
- OpenAI
- Pinecone
- Tailwind CSS
- Docker
Key decisions
- Separate the Chrome extension from the AI analysis service through a FastAPI backend.
- Use machine learning and NLP to analyze textual content.
- Use vector search to retrieve contextual reference material.
Architecture & approach
A Chrome extension handles content capture while a FastAPI backend performs the analysis. Machine learning and AI components process the content, while Pinecone vector search retrieves relevant reference material that can support the resulting credibility assessment.