Lina Weichbrodt is a generalist machine learning developer with a talent for prototyping data-driven products and getting them live. She covers the full lifecycle herself, from early design and implementation through A/B tests and day-to-day operations, working with a stack that spans AWS, Java, Python, and Spark. That end-to-end experience is rare, and it gives her a realistic view of what it actually takes to keep machine learning useful after the demo stage.

Her approach to machine learning operations is distinctly human-centered. She argues that models only create value when stakeholders trust them, so she focuses on practical tactics for buy-in: project intake checklists, pairing with business partners, demos over slideware, and turning stakeholder fears into concrete mitigations. On the technical side, she is equally hands-on with model monitoring, live test sets, feature drift detection, debugging, and ML incident response, from service levels and impact assessment to post-mortems and Five Whys root-cause analysis.

Lina has shared this experience with the wider community, including her talk Humans in the Loop at the DataTalks.Club 2021 Summer Marathon and a podcast deep dive on human-centered MLOps, model monitoring, and incident response, illustrated with a credit-scoring case study.