At Bennett University’s Quantum Computing Lab, I built an ML benchmarking pipeline for four QKD protocols. Feature engineering and parameter tuning reduced the key error rate by 18%.
I designed a model evaluation framework for QBER and Shannon entropy, with Matplotlib and Seaborn dashboards that cut analysis time by 3x. The framework detected eavesdropping above an 11% QBER threshold, and its reports were adopted as the lab’s standard format.
On my Signal Messenger Clone project, I developed a full-stack platform with Next.js, TypeScript, FastAPI, and WebSockets. It includes real-time messaging features, a normalized SQLite schema, and a Docker-based deployment with GitHub Actions and Jest.
I also created a customer churn prediction pipeline with NLP feature extraction and model evaluation, achieving 87% ROC-AUC. I’m exploring LLM APIs, RAG pipelines, and LangChain, and I’m available immediately.
