I'm Aurimas — an AI Engineer and writer from Vilnius, Lithuania, who came to engineering from chemistry. I build and ship LLM-powered systems: fraud/AML detection, RAG pipelines, data tooling, and AI features that work in production, not just in demos. I also taught Python and AI to hundreds of professionals across Lithuania and Latvia, and I write Molecule to Machine, a weekly newsletter on AI, science, and what actually ships.

name: Aurimas
located_in: Vilnius, Lithuania
role: AI Engineer (LLM, RAG, Fraud/AML, Data)
education: ["MChem, University of Edinburgh", "CS & AI Specialisations (Coursera/Databricks/Azure/AWS)"]
works_in: ["English", "Lithuanian", "French"]

core_stack:
  build: ["Python", "SQL", "JavaScript", "Flask", "Django", "REST APIs", "Docker", "Git", "CI/CD"]
  ai_ml: ["LLMs", "RAG", "prompt engineering", "LangChain", "fine-tuning", "PyTorch", "TensorFlow", "NLP"]
  data_cloud: ["AWS", "Azure", "Databricks", "Spark", "Airflow", "Pandas", "NumPy", "PostgreSQL", "MS SQL"]
  eval: ["evaluation harnesses", "guardrails", "monitoring", "human-in-the-loop design"]

interests: ["Applied AI", "Fraud/AML", "AI in healthcare & science", "Agentic systems", "Teaching AI"]

I own AI problems end-to-end — from the data and the prototype to a deployed, evaluated, human-in-the-loop system. A few things I've built to show how I work (live demos):

• ClearTrace Intelligence — AI fraud & AML investigation prototype with human-in-the-loop review.
• Aegis AI — LLM governance, observability & cost-tracking prototype.
• AI Gateway — product-led-growth onboarding prototype that tackles the cold-start problem.
• Agentic Mobile — B2B mobile AI-agent workspace prototype.

• Machine-Learning-Goodness — a curated collection of ML/AI resources, notebooks, and notes, with a focus on computer vision and ML in chemistry.
• CodeAcademy-AI-Course — full course materials and code from teaching AI with Python to adult learners.
• Python-Beginner-Course — Python teaching materials and exercises.
• Solutions-To-Problems — daily Python/SQL solutions to algorithm and interview problems.

I came to computing in chemistry research — molecular dynamics, bioinformatics, and scientific computing in Python and C++ — before moving fully into building AI systems. That scientific habit still shapes how I work: treat each initiative as a hypothesis, run a controlled rollout, read the metrics honestly, and iterate. Teaching Python and AI to non-technical people taught me the other half — that explaining a system clearly is most of what makes it useful.

I've been writing code in Python since 2010, and I'm a member of the Python Software Foundation, the Artificial Intelligence Association of Lithuania, the Association for the Advancement of Artificial Intelligence (AAAI), and the IEEE Computer Society.

• 🔭 Currently building AI prototypes and writing weekly — open to AI Engineer / Applied AI / GenAI Engineer roles across Lithuania, the Baltics, the Nordics, and EU-remote.
• 🌐 More at aurimas.io and Molecule to Machine.
• 💬 Happy to talk about LLM/RAG systems, fraud/AML, data engineering, or teaching AI to people who don't live in the details.