At Independent Power Transmission Operator (IPTO), I develop machine-learning workloads for the CAPE EU research project. I designed a real-time anomaly-detection pipeline combining a GDN deep-learning model with rule-based detection.
I also built and deployed an end-to-end workflow for data preparation, simulated streaming, model inference, anomaly scoring, and Grafana visualization across Ryax, IPTO on-premises infrastructure, and project hardware. For energy-demand forecasting, I developed a 24-hour workload using a dual-branch CNN–LSTM architecture.
Previously, at GenInvo, I developed an AI-assisted scanner for commercially sensitive and medical information across life-sciences websites and documents. My work also includes data-acquisition workflows, statistical analysis, and machine-learning experiments in internship and research settings.
