I care most about what happens after training: making models fast, reliable and useful in production.
An offline tool that removes secrets from text before I paste it into an AI assistant.
A team project that predicts the emotion in a short voice recording, built as our Le Wagon bootcamp final project.
A TU Berlin startup project exploring emotion recognition from infrared and face images, for therapy and wellness. Won Best Startup Idea in Berlin.
A side project to find good restaurants in Berlin by reading what reviews say about food, service and price.
A local tool that turns my digital activity into timesheets.
A personal system for collecting, scoring and following up on business ideas.
Freunde Messebau GmbH
Python · PyTorch · YOLO · OpenVINO · FastAPI · Docker · Linux
Humboldt-Universität zu Berlin
Le Wagon
Apple Kurfürstendamm
I'm Sina. I study computer science at Humboldt University in Berlin and work as a machine learning and software engineer, where I build and run computer vision systems on live video.
I got into machine learning through the Le Wagon Data Science & AI bootcamp in 2023. Since then I've worked on emotion recognition from speech and infrared images, and on a TU Berlin startup project that won Best Startup Idea in Berlin.
Outside of work, I build small tools I actually use, like an offline redactor that strips secrets from text before I share it with AI assistants.
Questions about a project, or work in computer vision and applied ML.