At Blinkin, I work on AI and ML systems, including a legal agreement analysis platform using BERT, PyTorch, and TensorFlow. I deployed LLMs and RAG and used OpenCV for OCR, reducing inference latency by 20%.
I also engineered an educational chatbot at Blinkin with Botpress, LLMs, RAG, and AWS ML, reducing its fallback rate by 15%. My work there includes production back ends and generative AI content tools.
Before Blinkin, I developed medical imaging models at MindTitan, including diagnostic systems for breast cancer and sarcoma. Earlier, at Eesti Post AS Tallinna filateelialTallinn, I developed predictive models and a speech-enabled analytics tool; I hold a Master of Artificial Intelligence and a Bachelor of Computer Science from Tallinn University of Technology.

