SoloSolve AI is an end-to-end system for transforming raw customer complaints into structured, actionable insights. It combines high-quality synthetic data generation, efficient QLoRA fine-tuning, and a 270M-parameter multi-task classifier that achieves near-perfect accuracy on Amazon Fashion complaint tasks.

End-to-end Machine Learning project using Random Forest, Flask web app, and AWS EC2 deployment.

A production-ready Machine Learning application that detects fake news in real-time using NLP and Logistic Regression. Features a decoupled training pipeline, a Flask REST API, and a web interface.