back to homepage
AI & Machine Learning
Applied AI/ML by Kacper Michalik—LLMs, RAG, evaluation, and MLOps shipped in real products.
Kacper Michalik’s AI and machine learning engineering focuses on shipping LLM-powered features that hold up in production: retrieval-augmented generation (RAG), prompt evaluation, guardrails, and MLOps around training, evaluation, and monitoring—built with Python, PyTorch, Pandas, and scikit-learn alongside full-stack product delivery.
Focus
- LLM product features with RAG, retrieval quality, and prompt evaluation
- Guardrails, observability, and gradual rollout for AI in production
- Classical ML literacy: splits, metrics, feature work, and overfitting awareness
- Pragmatic MLOps for small teams shipping user-facing AI
Stack
- Python
- PyTorch
- Pandas
- NumPy
- Matplotlib
- scikit-learn
- LLMs / RAG / prompt evaluation
- MLOps practices
In production
- O-1A Hub — founder; AI assistant with retrieval over U.S. O-1A visa guidance.
- AWS Certified AI Practitioner; writing on prompt engineering and model evaluation metrics (DZone).
Frequently Asked Questions
What AI and machine learning tools does Kacper Michalik use?
Kacper Michalik works with PyTorch for deep learning, Pandas and NumPy for data work, Matplotlib for visualization, and scikit-learn for classical ML. He ships LLM features with RAG, prompt evaluation, and MLOps practices.
Has Kacper Michalik shipped AI features in production products?
Yes. O-1A Hub is an AI assistant built around retrieval over U.S. visa guidance. He also holds the AWS Certified AI Practitioner credential.
Where can I read his AI/ML technical writing?
He has published on prompt engineering for generative AI and on model evaluation metrics on DZone, plus related ML topics such as training/validation/test data splits.
Related expertise
Related
