AI / Software Engineer - Private SME (Retail) s.r.o. (Private SME (Retail) s.r.o. AI and Web Development)
Duration: 10/2025 – present
Summary:
- Driving continuous upskilling with practical delivery focusing on LLM-powered automation, backend engineering, and developer workflows
- Prototyping AI-assisted content and operations tooling and developing a modern web presence for a boutique brand including a website MVP and evaluating e-commerce integration options
Responsibilities:
- Developing AI-assisted tools and backend systems using modern technologies.
- Prototyping and implementing LLM-powered automation solutions.
- Creating and maintaining a web presence for the company’s boutique brand.
- Continuous professional development in AI and machine learning fields.
Technologies: Python · FastAPI · LLMs · LangChain/LangGraph · Git · GitHub · GitHub Copilot · Docker · PyTest · Markdown · AI-assisted development (PyCharm/Cursor)
AI / Software Engineer - Raiffeisenbank (client) (Change Impact Analysis MVP with AI Agents)
Duration: 5/2026 – 7/2026
Summary:
- Developed an AI agent-based MVP for change impact analysis, including a User Guide Agent (UGA) orchestrating deeper analysis agents
- The project aimed to provide a user-facing orchestration system for analyzing change impacts using AI agents
Responsibilities:
- Built Python and LangGraph-based AI agents for change impact analysis.
- Implemented new UGA endpoint with AWS Lambda integration.
- Designed and refined system prompts for orchestration and user guidance.
- Applied AI-assisted development workflow including code review and issue tracking.
Technologies: Python · LangGraph · LLM Agents · Prompt Engineering · AWS Lambda · AWS/AWS CLI · GitHub Enterprise · GitHub Copilot CLI · Git · JIRA · AI-assisted development · Code Review
Senior AI Engineer - PricewaterhouseCoopers (client) (Audit-Ready Financial Statements Automation)
Duration: 8/2025 – 10/2025
Summary: Automated the generation of audit-ready financial statements by replacing manual Excel to Word workflows with a reliable, auditable GCaaS pipeline for the TLP team, improving runtime, reliability, and reducing manual toil.
Responsibilities:
- Mapped multi-sheet Excel datasets into predefined Word templates.
- Built version-controlled LangFlow flows synchronized with Python sources.
- Enforced Python 3.12 and pinned library versions for stability.
- Improved runtime and reliability through testing and monitoring.
- Designed prompt-and-tooling framework with evaluation loops and regression tests.
- Built document extraction pipeline with structured outputs and reliability constraints.
Technologies: Python 3.12 · LangFlow · GitHub Copilot (GPT 4.1) · Git · GitHub (code reviews, tags, releases) · PyTests · Markdown · SCRUM (Kanban) · vibe-code-centric development (PyCharm)
Machine Learning Engineer - Rohlı́k Group (client) (Supply Chain & Demand Forecasting Service)
Duration: 5/2025 – 7/2025
Summary: Developed a new FastAPI service with custom endpoints for monitoring, evaluation hooks, and automated test coverage to support reproducible results in supply chain and demand forecasting.
Responsibilities:
- Developed FastAPI service with custom endpoints for monitoring and evaluation.
- Implemented automated test coverage for reproducibility.
- Optimized performance and resolved issues using Cursor AI and other tools.
- Obtained Snowflake SQL and Keboola Data Academy certifications during the engagement.
Technologies: Python · Cursor.ai · FastAPI · pytest · Git/Gitlab · Claude Code · Code Review · Google Gemini · Anthropic Claude · Bash · Polars · Google Workspace
Member of Technical Staff - Pure Storage (client) (Digital Experience Prompt Engineering)
Duration: 10/2024 – 1/2025
Summary: Worked on prompt engineering and prompt meta-engineering to enhance digital experience projects, utilizing advanced AI models and tools until project closure due to internal restructuring.
Responsibilities:
- Performed prompt engineering and meta-engineering for AI models.
- Utilized ChatGPT Enterprise and GPT models for project tasks.
- Collaborated using GitHub, Kubernetes, and other DevOps tools.
Technologies: ChatGPT Enterprise · GPT 4o & o1 · Github Copilot · Python · FastAPI · PyTest · Pydantic · Github · Kubernetes · Watchdog · Elastic · Spinnaker · Google Suite
AI Specialist & Prompt Engineer - Aimtec, Division of Integration (Information Extraction from Unstructured Documents)
Duration: 2/2024 – 8/2024
Summary: Designed and implemented an LLM-powered document extraction pipeline with structured outputs and reliability constraints using AWS Bedrock, focusing on prompt engineering and AI-assisted development.
Responsibilities:
- Designed prompt-and-tooling framework with evaluation loops for document processing.
- Built document extraction pipeline using AWS Bedrock.
- Applied AI-assisted development techniques with Copilot and GPT models.
Technologies: prompt engineering · prompt meta-engineering · Claude {3, 3.5} Sonnet · AWS Bedrock · Python · bash · vim(diff) · Onenote
AI Development Engineer – Autonomous Driving projects - Porsche Engineering (ADAS Team Autonomous Driving Projects)
Duration: 6/2023 – 10/2023
Summary: Worked on machine learning and computer vision projects related to autonomous driving, focusing on data science and data imputation techniques.
Responsibilities:
- Developed machine learning models for autonomous driving applications.
- Applied computer vision techniques and data imputation.
Technologies: ML · computer vision · Python · Transformers · data science · data imputation
AI Researcher at Computer Vision Research Group - Technical University of Munich, Chair of Remote Sensing (Deep-learning Hybrid Uncertainty-aware Modelling with Earth Observation Data)
Duration: 3/2022 – 4/2023
Summary: Conducted research on deep learning models for semantic segmentation of Earth observation data using Monte Carlo Dropout on UNet architecture, contributing to uncertainty-aware modelling.
Responsibilities:
- Developed and evaluated deep learning models for semantic segmentation.
- Applied Monte Carlo Dropout techniques for uncertainty estimation.
- Managed experiments using PyTorch, Weights & Biases, and GPU computing.
Technologies: PyTorch · Weights & Biases · GPU Computing · GitLab
AI Specialist - Volkswagen Group (Digiteq) (Video Perception for Autonomous Driving / Advanced Driver-Assistance Systems)
Duration: 12/2021 – 2/2022
Summary: Worked on video perception and core network architectures optimization for autonomous driving and advanced driver-assistance systems, including network quantization and TensorRT optimization.
Responsibilities:
- Developed and optimized core network architectures for video perception.
- Implemented network quantization and TensorRT optimizations.
Technologies: TensorRT · network quantization
Research Scientist (part-time internship) - IBM, Watson Research (Natural Language Processing and Mobile Computing Research)
Duration: 5/2013–5/2014 & 10/2014–3/2015
Summary: Conducted research projects focused on improving speech recognition accuracy and processing crowd-annotated datasets for ellipsis resolution in natural language processing.
Responsibilities:
- Improved accuracy of speech recognition for name utterances.
- Processed and verified crowd-annotated datasets for ellipsis resolution.
Technical Studentship Programme - CERN, Intel & LHCb Online Computing (Realtime Applications for LHCb Detector Online DAQ System)
Duration: 4/2015–3/2016
Summary: Developed realtime applications for the LHCb detector's online data acquisition system, improving event sorting and implementing buffered read techniques.
Responsibilities:
- Improved event sorting using OpenMP parallelization and loop tiling in C++.
- Implemented buffered read for LHCb DAQ system in C.
Technologies: C++ · C · OpenMP
Summer Studentship Programme - CERN, Data Storage Group (Tape Storage and Data Protection Project)
Duration: 6/2014–9/2014
Summary: Worked on tape storage and data protection by increasing speed of CRC-enabled tape write through multithreaded CRC error correction implementation.
Responsibilities:
- Implemented multithreaded CRC error correction using C pthreads.
Technologies: C · pthreads
Open Source Software Engineer - H2O.ai (PCA Functionality Enhancement)
Duration: 4/2017–9/2017
Summary: Enhanced the speed of PCA functionality by integrating and benchmarking alternative Java implementations of SVD decomposition, including unit and performance tests.
Responsibilities:
- Integrated and benchmarked Java SVD implementations for PCA speed improvement.
- Developed unit and performance tests.
Technologies: Java
AI Researcher - Artificial Intelligence Center, Czech Technical University (Complex Imperfect-Information Game Strategies using Deep Learning)
Duration: 10/2017–1/2019
Summary: Researched imperfect-information game theory and deep learning, implementing tree search enhanced with neural networks using TensorFlow, resulting in a journal publication.
Responsibilities:
- Developed deep learning models for imperfect-information games.
- Implemented tree search enhanced with neural networks.
- Published research in AI journals.
Technologies: TensorFlow
Researcher - Imperial College London, Faculty of Natural Sciences (Deep Learning for Evolutionary Inferences from Large Genomic Data)
Duration: 4/2019 – 10/2020
Summary: Applied machine learning techniques to population genomics data of malaria mosquitoes for biological and evolutionary inferences such as migrations and selection strength.
Responsibilities:
- Analyzed large genomic datasets using machine learning.
- Derived biological and evolutionary insights from data.
Technologies: Pandas · NumPy
Upstaffer since September, 2026
