Work & Projects
A collection of my work in machine learning, security research, and educational tools.
A live, source-grounded portfolio assistant that retrieves verified evidence before answering visitors.
A live multi-workspace tracker with an auditable XP ledger, analytics, AI weekly reviews, and long-term progress tools.
A live hybrid recommender spanning reproducible ML experiments, a Flask API, a React client, and AWS deployment.
An 11-experiment LLM engineering study from a 112K-parameter character model through Qwen LoRA and grounded retrieval.
A live regional prediction platform with player accounts, public leaderboards, tournament views, and a private command center.
Scheduled LLM agents authenticate as players, reason over eligible fixtures, and submit structured predictions to the live game.
An AQI forecasting prototype combining satellite observations, ground measurements and weather context.
A black-box evaluation framework for jailbreak attempts, refusal behavior, latency, token use, and model failure modes.
A four-repository implementation series covering video captioning, spoken question answering, and generative adversarial networks.
Computer vision system for vehicle license plate detection and text extraction.
Real-time monitoring system using computer vision for safety compliance.
Research & Experimentation
Hands-on experiments, security research, and applied ML case studies — work that lives at the boundary of exploration and engineering.
