Skip to content

About Me

Sławomir Strzelec

Hello! 👋 My name is Sławomir Strzelec. I'm an AI/ML Engineer with a research background in organic chemistry, nanotechnology, and spectroscopy — and hands-on engineering practice building AI systems end to end, from retrieval-augmented generation pipelines to tested, publicly deployed REST APIs.

I combine a rigorous scientific foundation with modern software engineering practices: not just prototyping ML models, but packaging them as tested, containerized, and publicly deployed systems. My flagship project — a Carbon Nanotubes RAG System — went through six deliberate development phases: correct retrieval math, a validated REST API, structured logging, Docker, CI/CD, and live deployment. It's the clearest demonstration of how I work.

What I Focus On

  • Retrieval-Augmented Generation (RAG) and agentic AI systems
  • Building and deploying REST APIs (FastAPI) around ML/AI pipelines
  • LLM integration, prompt engineering, and AI evaluation (RAGAs)
  • Computer Vision and image-based analysis
  • Data processing, analysis, and modeling in Python
  • Scalable data engineering and analytics pipelines
  • Scientific computing and analytical reasoning

Carbon Nanotubes RAG System

A Streamlit prototype rebuilt into a deployed, tested system: hybrid dense+BM25 retrieval, a validated REST API, 31 automated tests (CI via GitHub Actions), Docker containers, and a live public deployment.

  • Faithfulness ≈ 0.80–0.82 (RAGAs evaluation, 14-question domain set)
  • 31/31 tests passing, enforced automatically on every push
  • 🔌 Live REST API · 💻 Live Demo · 📦 Repository

Skills

Core (used in my projects daily)

  • Python, FastAPI, REST API design, pytest
  • LLM & agentic AI: RAG, LangChain, LangGraph, OpenAI API, prompt engineering, RAGAs evaluation
  • Vector search: FAISS, embeddings, hybrid (dense + BM25) retrieval
  • ML / DL: scikit-learn, PyTorch, XGBoost, model evaluation
  • Engineering: Docker, GitHub Actions (CI/CD), Git, structured logging
  • SQL, PostgreSQL

Working knowledge

  • Computer vision: OpenCV, CNN-based models
  • Chemoinformatics: RDKit, molecular descriptors
  • Streamlit, EDA, statistics, data visualization (Plotly, Looker Studio)
  • Spectroscopy and scientific computing: NMR, Raman, UV–Vis, SciFinder, Reaxys

Familiar

  • PySpark, BigQuery, AWS SageMaker, Linux CLI

Also: grant writing, teaching and mentoring, science communication.

My Background

🎓 Education

  • PhD (in progress) – AGH University of Krakow Interdisciplinary Environmental PhD Studies (physics, chemistry, biophysics, nanotechnology)

  • MSc in Chemistry – Jagiellonian University Modern organic synthesis & physicochemistry

  • BSc in Chemistry – Jagiellonian University Synthesis and catalytic studies of chiral half-salen complexes

🔬 Scientific & Industry Experience

  • Data Engineer – HSBC (Apr 2026 – May 2026) ETL/ELT pipelines, BigQuery, Looker Studio dashboards

  • Scientist II – Selvita (2024–2025) Data analysis and experimental work in early-stage drug discovery; compound design, synthesis, and characterization

  • Internship – University of Turin (2023) Organic synthesis and reaction mechanism studies

  • UNISON Project (2018–2019) Authored and defended a national R&D funding application before an NCBR expert panel — application approved

🧪 Scientific Interests

Organic chemistry, nanotechnology, spectroscopy, machine learning in science, quantum physics, photovoltaics, drug discovery, targeted cancer therapies, energy storage.

📚 Publications & Awards

  • Scientific publication (2025): "Study of the Influence of Multi-Walled Carbon Nanotubes Functionalised with Nickel Ions on the Functioning of Red Blood Cells" — Acta Physica Polonica A

  • Rector's 1st-degree award for pioneering work in micro and nanotechnologies in biophysics

  • Winner of a Jagiellonian University chemistry competition

My Projects

My projects reflect both my scientific background and my engineering practice: RAG and agentic AI systems, computer vision for biomedical imaging, machine learning models for chemical property prediction, and end-to-end ML applications with deployment — not just notebooks, but tested and documented systems.

→ See all projects

Contact

I'm open to new challenges, collaborations, and opportunities in AI/ML engineering, agentic AI, and data science. 🚀