CV & Experience

Profile

Work

• 2026.10 - Present

  Software Engineer

  Google

  Software Engineer on the Agent Development Lifecycle team at Google.

• 2025.09 - 2026.09

  Software Engineer

  Box

  Worked with the team to create a workflow platform with its own expression language.

  • Workflow platform development
  • Platform with its own expression language
  • Java + GraphQL backend development
  • GCP cloud infrastructure

• 2025.05 - 2025.08

  Data Scientist

  Capgemini

  Designed and implemented fullstack agentic application for automating claim management process using LangGraph leveraging Azure cloud.

  • Fullstack agentic application development
  • LangGraph implementation
  • Azure cloud deployment
  • Claim management automation

• 2024.10 - 2025.06

  ML Software Engineer

  Mi2.AI

  Designed fullstack webapp for cancer prediction models, with multistep inference logic using RabbitMQ. Implemented Java Spring backend and React TS frontend. Deployed and tested the system on GPU powered kubernetes cluster.

  • Fullstack webapp development
  • Java Spring backend
  • React TypeScript frontend
  • RabbitMQ message queue
  • GPU powered Kubernetes deployment
  • Cancer prediction ML models

• 2023.02 - 2025.02

  Co-Founder / CTO

  APPI

  Designed and created knowledge graph based RAG system, using LlamaIndex, Langchain and Pinecone, to parse, retrieve and evaluate context from documents. Deployed with FastAPI on Azure.

  • Knowledge graph based RAG system
  • LlamaIndex and LangChain integration
  • Pinecone vector database
  • FastAPI deployment on Azure

• 2022.07 - 2025.07

  Working Student / ML Software Engineer / Software Developer

  Nokia

  Designed, created and deployed reinforced learning (DDQN), PyTorch deep learning and tuned open source time series prediction models for kubernetes pod autoscaling. Designed and implemented automatic, configurable training pipelines and MIFlow based model registries on remote GPU powered remote lab. Configured Jenkins pipelines for CI/CD and Helm charts for Openshift labs.

  • Reinforcement learning (DDQN) models
  • PyTorch deep learning
  • Time series prediction models
  • Kubernetes pod autoscaling
  • MIFlow model registry
  • Jenkins CI/CD pipelines
  • Helm charts for Openshift

Education

• 2026 - Present

  Warsaw, Poland

  Master's Degree

  Warsaw University of Technology

  Data Science - Faculty of Mathematics and Information Science

• 2022.10 - 2026.02

  Warsaw, Poland

  Bachelor's Degree (completed)

  Warsaw University of Technology

  Data Science - Faculty of Mathematics and Information Science

  • Calculus II
  • Theory of Probability
  • Statistics
  • Algorithms
  • ML
  • Data visualization
  • Databases

Projects

• Mi-Crow

  Co-developed a Python library for mechanistic interpretability of large language models as my engineering thesis at Warsaw University of Technology, with Hubert Kowalski.

  • Sparse autoencoder training (TopK and L1)
  • Activation analysis and hooks
  • Concept discovery and model steering
  • Unified Hugging Face model interface
  • Python, PyTorch, and Transformers
• 2024.10 - 2025.06

  xLungs

  Created a Java + React fullstack application that performed inference for the xLungs medical imaging model, supporting CT image processing, organ segmentation, and report generation for radiologists.

  • Java Spring backend
  • React TypeScript frontend
  • Model inference integration
  • Multistep inference pipeline with RabbitMQ
  • GPU-powered Kubernetes deployment
  • CT image processing and report generation
• 2023.02 - 2025.02

  APPI Synapse

  An internal tool for information retrieval using RAG techniques, developed for the architecture and construction industries.

  • Integrated React, FastAPI, and MongoDB
  • Leveraged LlamaIndex and LangChain for document processing
• 2023.02 - 2025.02

  APPI Marketplace

  AI agent platform designed for technical and engineering documents, with a focus on advanced, reliable retrieval and document-grounded responses.

  • Knowledge graph-based RAG for document retrieval
  • Context retrieval and evaluation for AI agents
  • LlamaIndex, LangChain, and Pinecone
  • FastAPI backend deployed on Azure
• Terrain Segmentation Models

  Developed terrain image-processing models as a member of the Project Trailblazer CanSat team; source notebooks and example outputs are available in the project case study.

Awards

• CanSat Finalist

  CanSat

  Finalist in the CanSat competition.

Skills

Languages

Interests

Volunteer

• Mentor

  BEST Hacking League

  Mentored teams on BEST Hacking League hackathon in the field of AI.

  • AI hackathon mentoring
  • Team guidance
  • Technical support
• Speaker

  No Code Days & University AI Clubs

  Delivered lectures about fine-tuning and embeddings at No Code Days and university AI clubs.

  • Fine-tuning lectures
  • Embeddings presentations
  • AI education
• IT Volunteer / Web Developer

  Co.Hana Brain Foundation

  Developed and maintain the foundation's website, providing IT support for a non-profit organization promoting brain health, prevention of neurodegenerative diseases, and Alzheimer's research.

  • Website development and maintenance
  • IT support
  • Non-profit technology solutions
  • Brain health advocacy