
I'm Marius Ailinca, an AI engineer and Rust developer based in Bucharest. I've spent 12+ years building software
across AI R&D, real-time systems, and full-stack development.
My interest in physics started early — I competed in Romania's National Physics Olympiad two years running in high
school. That mathematical foundation is what I lean on now for AEGIS's Kalman filter design, sensor noise models,
and trajectory prediction. I later studied AI at Heriot-Watt University (MSc with Distinction, GCHQ-certified
programme) and Computer Science at Vasile Alecsandri University (BSc with Distinction). I've published research on
neural network topologies for robotics applications.
By day I work as an AI Engineer at Stefanini EMEA, building enterprise ML systems — HybridRAG architectures,
multi-agent platforms, and MCP servers. I'm also the author of codescout, an open-source Rust MCP server that
gives AI coding agents IDE-grade code intelligence.
For AEGIS, I built the full pipeline solo: the Rust IMM/Kalman tracker (50–200x faster than Python), the
multi-modal ONNX drone classifier, the trajectory predictor, the NATS microservice architecture, and the
React/Mapbox dashboard. All sensor parameters are calibrated against Ukrainian conflict field data — the physics
background helps when you're tuning process noise covariances and radar range equations.