Luka M.

I'm an AI Automation Engineer based in Italy, specializing in building systems that monitor data and automate processes on their own — from anomaly detection to LLM-based workflows. What I do: Anomaly detection and data monitoring systems. I build unsupervised pipelines (K-Means, Z-score) that catch operational problems before they become critical. For example: two anomaly detection engines for a 500+ device IoT fleet across Germany and Italy, reaching around 95.7% precision (projects public on GitHub). AI-driven workflow automation. Custom Python scripts, LLM API integration (OpenAI, Anthropic), pipelines that connect data, tools, and models with no manual intervention needed. Data engineering. I turn messy, raw datasets into analysis-ready outputs, with a strong focus on precision and optimization (Python: Pandas, NumPy). Background: I'm completing a Higher Technical Diploma in AI and Data Analytics, with hands-on experience evaluating and training language models (RLHF). That work gave me a deep understanding of how LLMs actually behave, which helps both when building reliable automations and when debugging things that go wrong. Native speaker of Italian and Serbian, professional English. Send me a quick description of your current workflow or problem. I'll come back with a concrete idea of where to start.