AI Engineer — Poznań / Szczecin
Sebastian Brzustowicz
I design and ship autonomous AI systems — AI Agents, Prompt Engineering and Azure Solution Architecture — building secure, scalable agentic workflows that bridge cutting-edge models with robust enterprise infrastructure.
About
My focus sits at the intersection of artificial intelligence and production engineering. I build agentic systems that actually work in the enterprise.
From multi-agent orchestration with LangGraph and LangChain to vector search with Qdrant and Chroma, I connect state-of-the-art language models with robust cloud architecture on Azure — turning prototypes into dependable, secure products.
With a background spanning software engineering, robotics and computer vision, I bring a full-stack perspective to every AI project: the model, the prompt strategy, the database, the API layer and the deployment pipeline.
Skills
Cloud & Architecture
- Azure
- Solution Architecture
- AI Security
AI & Agent Systems
- AI Agents
- Prompt Engineering
- LangGraph
- LangChain
Model Engineering & MLOps
- PyTorch
- MLOps
- MLflow
Languages & Backend
- Python
- SQL
- TypeScript
- FastAPI
- Bash
Databases & Vector Search
- Qdrant
- Chroma
- PostgreSQL
DevOps & Infrastructure
- Docker
- CI/CD
- Airflow
Experience
AI Engineer — Adamus S.A., Szczecin
Architected and deployed an enterprise multi-agent system on Azure, incorporating prompt engineering to automate quotation.
Software Engineer — ITTI Sp. z o.o., Poznań
Designed scalable backend architecture and APIs for spatial and simulation platforms using Docker and PostgreSQL.
Robotic & PLC Programmer — InPro engineering, Poznań & Bremen
Programmed a collaborative robot (cobot) to synchronize movements with human motion captured via computer vision.
Selected work
Robot Sumo RL
Advanced Reinforcement Learning (SAC/PPO/A2C) for autonomous Robot Sumo combat.
Python · PyTorch · RL 02Rocket Control
High-fidelity simulation for rocket flight dynamics and control — ideal for ML/RL training.
Python · Simulation · GNC 03Rubik2x2 DQN
Multiparadigm DQN + Imitation Learning project solving a 2×2 Rubik's cube.
Python · PyTorch · MLflow 04Collision Detection AI
ML model for collision detection trained on accelerometer sensor data with PyTorch.
Python · PyTorch · Sensors 05Wide-Gemini
Browser extension that enhances the Gemini experience with a cleaner reading layout.
JavaScript · Extension 06Quadcopter Control System
Quadcopter control system simulation based on a linear controller with visualisation.
MATLAB · Simulink · ControlEducation
Master of Engineering
Poznan University of Technology — Automation, Robotics and Electrical Engineering
Engineer's degree
Poznan University of Technology — Computer Science
- Harvard University — Data Science: Building ML Models
- EITCA — Advanced Reinforcement Learning
- EITCA — Machine Learning with Python
- EITCA — Deep Learning with Python and PyTorch
- Polish — native (C2)
- English — fluent (C1)
- German — basic (A2)