AI-powered misinformation detection
Verity
An end-to-end NLP pipeline — model training, API deployment, and a production-ready frontend — for scoring the credibility of news text.
The challenge
Most misinformation-detection demos stop at a Jupyter notebook. The goal was to ship something a real user could open in a browser and get an answer from in under a second.
The approach
Text is cleaned and vectorized with TF-IDF, then scored by a logistic regression model trained on labeled news data. The model is served behind a Flask API with input validation, and a static frontend calls the API directly — no unnecessary middle layer.
The result
A working public tool: paste an article, get a credibility read. Built end to end, deployed, and maintained as a real service rather than a one-off script.
Technology
- Python
- scikit-learn
- Flask
- Render
- Netlify