I start from the problem, build the system, and measure whether it solved it. Looking for a data science or machine learning internship from February 2027.
A voice assistant for a reception robot that answers only from verified information.
Real-time recognition of Dutch Sign Language fingerspelling, with live monitoring.
Team project; I built the operations dashboard.
How internet access relates to school completion across ten South American countries.
A face-shape classifier for eyewear, evaluated against human agreement.
Forecasting the NASDAQ-100 opening gap from overnight Asian and European markets.
Team project, which I led.
Emotion recognition in Bulgarian speech, from a YouTube link to a labelled transcript.
Team project; I built the LSTM track and co-built the transformer.
A simulated vision-to-control pipeline for inoculating plant root tips.
How well a modern object detector performs on historical street photos.
Return and warranty decisions grounded in a company's written policy, using RAG.
Team of two; I built the RAG part.
July 2026 to present
Breda, Netherlands
Breda University of Applied Sciences
Mentor: Margot Neggers
Read the case study →I'm happiest in the messy middle of a project: the part where an idea has to survive imperfect data, honest feedback, and a deadline.
I came into AI fascinated by what becomes possible with enough data and the technology now emerging around it. But I've always been just as certain about the other half: that whatever I build has to actually matter. That combination of curiosity about what's possible and a stubborn focus on whether it works for the person it's built for is what keeps pulling me deeper into the field.
That curiosity is also what moved me from Bulgaria to the Netherlands, to study in a degree built around real projects and real clients. That is where my breadth across language, vision, human behaviour, robotics and cloud systems comes from, and it lets me get productive fast in whatever a problem actually needs. The throughline never changes: I like to build, and I start from the person it's for.
I'm a data science & AI student now looking for an internship where I can keep doing exactly that: building systems that solve a clear problem and hold up when tested.
TensorFlow · OpenCV · YOLOv11
Hugging Face · ChromaDB · LangChain
Stable-Baselines3 · PyBullet · SAC
Docker · Azure · FastAPI · logging
If you're building something where the model has to survive real users and a real deadline, I'd love to hear about it. The fastest way to reach me is a direct email. I read every one.
Email me directly ry.kichukova@gmail.com write an email ↗