Emilio Oldenziel

About

Computing scientist and tinkerer specialized in MLOps and deep learning. I like building things, understanding how they work, and occasionally writing about both.

Publications

2021 Provident Detection of Vehicles at Night IEEE A camera-based method for detecting oncoming vehicles at night via headlight beam features, aimed at improving reaction time in ADAS applications.
2018 A comparison of ensemble methods: AdaBoost and random forests Best Paper Winner of the Student Colloquium Best Paper Award 2018 (University of Groningen). See page 5 and 65 of the proceedings.

Blog

2026 — Apr Unix cli is all you need: a minimal LLM agent harness, introducing Wiekly A pure Python agent harness where the model writes Unix pipelines as tool calls and calls back itself via curl, with no subprocess, no eval, and no sandboxing needed.

Writing & Media

2025 — Aug · Podcast The Next S-Curve – GenAI in contractmanagement: van maanden naar minuten How generative AI and agentic AI can compress contract management workflows from months to minutes while keeping humans in the loop.
2024 — Sep · Article Toepassing van geavanceerde machine learning op mission-critical operaties A case study with a European railway company on scaling ML research into production systems using MLOps and human-in-the-loop approaches.
2024 — Mar · Whitepaper Winning the data analytics race of Transport & Logistics A business-centric playbook for data analytics in transport and logistics, arguing for shifting data ownership from IT to business units.
2021 — Jan · Article Provident Detection of Vehicles at Night Interview on the research behind camera-based nighttime vehicle detection and its application in advanced driver-assistance systems.