AI/ML Engineer Intern at Datarabbit

Worked on an AI-powered app to streamline lead analysis and personalization for cold outbound messaging. Reduced lead cost from $0.02 to $0.005 by implementing new LLM logic in Lead Analysis. Migrated LLM from 4o-mini to 5-nano and tuned AWS infra limits via Terraform, increasing leads per request from 3 to 30 in Lead Finder and Mail Generation microservices. Also contributed to a Voice Assistant Demo for patient follow-ups at Poznań hospital.

Data Scientist Intern at Visimind

Refactored the real-time LIDAR data processing system, modernized library versions, and achieved a 25% time reduction in background filtering (36.3 ms to 27.4 ms via Numba). Developed a Python tool to convert raw LIDAR data from multiple models (Velodyne, Ouster) into standardized reports.

Board Member at Czarna Magia – Student AI Society

Participated in projects as part of the team, performed code reviews, prepared technical workshop materials, and managed weekly social media communication.