Leonardo Schwarz · Software & AI Engineer

Let’s evaluate and bring it to production.

What I build

Data pipelinesIngestion, SQL analytics and reproducible processing for scientific and business data.
Backends and internal toolsAPIs, dashboards and interactive analysis UIs on top of your data and systems.
Software deliveryContainers, CI/CD, automated testing and reproducible environments.
ML engineeringDeveloping and integrating models into larger systems.
LLM integrationAI coding agents and prompt workflows set up for the work you actually do.

Evals, in practice

I turn “will it work?” into something you can test: tasks from your real cases, data to run them on, and a way to score the output. Before anything ships, the test has the answer: numbers and a failure list. Run it again whenever something changes. And when it finds a problem, I help fix it.

Now

I’ve been building ML and research software in Zürich since 2021. At a medical imaging company I validated deep learning models for 3D MRI diagnostics, and the evaluation results went into a regulatory approval submission. Most recently: data pipelines and analysis tooling at the Functional Genomics Center Zürich (ETH Zürich).

Something that should work, and nobody’s sure it does? Something that doesn’t, and it needs to? Write me and we’ll take a look.