I work on AI and ML that are reliable and efficient. My experience spans finance and healthcare, including tuning FinBERT for sentiment and building a variational quantum model for cancer classification. I use both classical optimization and quantum methods such as QAOA and VQE, and I share clean code with clear steps to reproduce results. For a PhD, I want to study optimization for large models, robustness and privacy, and ML systems that run on real hardware in the cloud, on GPUs, and at the edge and in IoT. I am open to NLP, time series, tabular, and multimodal data, and to work that connects ML with security and data systems, with the goal of methods that are simple to use, grounded in theory, and ready for deployment in health, finance, and beyond.
Manuscript available upon request.
Programming & Platforms: Python (PyTorch, scikit-learn, NumPy, Pandas) · Qiskit · SQL · Git/Linux · MLflow/W&B · Docker (basic) · GPUs · cloud · edge/IoT
Trustworthy ML & Stats: Transformers · GNNs · multimodal fusion · robust training · calibration & uncertainty · ablations/error analysis
Quantum & Optimization: VQE/QAOA · hybrid C-Q workflows · QUBO/Ising mappings · quantum-guided HPO
Data Eng & Reproducibility: SQL · ETL · schema design · MLflow/W&B · env pinning · unit tests · CI/CD (Actions) · DVC
Specialized Skills: Privacy (HE, differential privacy) · Clinical NLP (HF fine-tuning, ASR) · Visualization (LaTeX, Matplotlib, dashboards)
Domains: Computational genomics · clinical decision support · financial sentiment