AI & LLM applications
Building, shipping, and evaluating LLM-powered applications. From prompt engineering to production-grade agent systems.
Menu
Speaking.
I bring practical, experience-driven talks to conferences, meetups, and corporate events. 24 talks shipped to date — from AI Engineer Europe to dbt Coalesce to EuroPython. Whether your event is 30 people or 3,000.
What I talk about.
Evals & testing for AI
How to evaluate AI systems rigorously — synthetic data, automated testing, and continuous evaluation pipelines.
Data engineering
Modern data stacks, dbt, DuckDB, and bringing data engineering practices to small teams and startups.
Vibe coding
Running multiple AI coding agents in parallel; live demos of AI-first development workflows.
Python patterns
Pydantic, type safety, and practical Python patterns for the LLM era.
Your topic?
Tell me about your audience and theme; I'll craft a talk that fits.
talk to me ↗Catch me next.
2026Evaluace v praxiCERNA.AI (Ostrava, CZ)Všechny evals frameworky jsou tak trošku na nic. Jak je přesto využít?Moderní stack #7 (Ostrava, CZ)Jak poznat, že je AI co k čemu?Dobrodata (Prague, CZ)
Upcoming talk at CERNA.AI festival.15 Sep 2026 · register ↗
Eval frameworks all promise more rigor than they deliver — but writing them off entirely is the wrong move. This talk covers what they actually get right, where they quietly mislead, and how to use them anyway to learn something true about your system.30 Sep 2026 · register ↗
This talk deliberately isn't about what AI can do right now — that shifts again by Q3 — but about a harder problem: your own sense of productivity is unreliable, and a "good" result in one area says nothing about the one beside it. You'll leave with one question to ask constantly ("is this actually working for me?") and one concrete habit for answering it — finding your own pelican.25 Nov 2026 · register ↗
Past engagements.
Hover any entry; some have abstracts. Featured talks are markedin flame.
2026RAG a evalyAI Monday Jihlava #7 (Jihlava, CZ)Choose Your Ground Truth: A Field Guide to Synthetic Data for EvalsEvals.cz Meetup #3 (Prague, CZ)Financial modelling in OpenClaw & safely deploying itOpenClaw Demo Night (Prague, CZ)Cognitive Exhaust Fumes: What Read-Only AI Sees That You Can'tai.engineer/europe (Online)The Eval Flywheel: From "Works on My Laptop" to Systematic QualityGenAI Meetup (Prague, CZ)Evals, Benchmarks, and Guardrails: A Pythonista's Guide to Not Mixing Them UpPyData (Prague, CZ)When (& How) to Start Writing EvalsEvals.cz Meetup #1 (Prague, CZ)Simultaneous Vibe CodingTopMonks Caffé (Prague, Loft N8)2025Jak se učit s AIZvídavá nadílka 2025 (Prague, Kino Pilotů)What Makes Us Human?Data Day 2025 (Prague, CZ)Getting buy-in for data infrastructure setupdbt Coalesce (Las Vegas, NV, USA)Simultaneous Vibe CodingAI Bootcamp 2025 (Chocerady, CZ)Evals & Synthetic DataHacker Camp 2025 (Rožmitál pod Třemšínem, CZ)Pydantic, Everywhere, All at OnceEuroPython (Prague, CZ)Debate ConciergeAI Tinkerers (Prague, CZ)2024Multi-Agent FrameworksPyData Prague, Rapid7 offices, CZLLM & TDD? Maybe! (& AutoGen & DeepEval)pyvo Praha (Prague, CZ)2023Impromptu Talk at Data Mesh @ World Data Congress 2023World Data Congress 2023, Prague, CZWebinář: Umělá inteligence a její hrozbyPrague, CZ2022dbt and MDS in small-batch academic research: a working exampledbt Coalesce (New Orleans, LA, USA)Data Modeling as the First Data HireSTRV, Prague, CZNeuroscience to MDS, or: On Data As ExhaustSummer Community Days (Online)2017Data Curation WorkshopCharlottesville, VA, USASIPS Lightning TalkCharlottesville, VA, USA
"RAG is dead" is the take in every other thread in 2026 — and it's wrong: naive retrieval-augmented generation is still a sensible default, beaten only in some cases, and measurement is the only way to know if you're one of them. This talk walks the retrieval pipeline end to end, then turns to the part that matters — telling whether your RAG actually works, with ground truth, retrieval metrics, RAGAS, LLM-as-judge, and error analysis feeding an eval flywheel.Jun 2026
You need an eval set but don't have a hundred real production failures to build it from, so you reach for synthetic data — and most first attempts quietly produce garbage. A field guide to the techniques that actually work, from real-incident seeds to personas to RAG-grounded generation, with one throughline: synthetic data needs its own eval, so choose your technique backwards from the eval you want.Jun 2026
Using OpenClaw to build a Bayesian buy-vs-rent model for Prague real estate, and how to deploy something like that without setting your money on fire.Apr 2026
What happens when AI systems passively observe information without modifying it? Exploring the patterns and insights that read-only AI reveals — the cognitive byproducts humans overlook.Apr 2026
Most teams shipping GenAI products have no evaluation system. They have vibes, a few saved prompts, and hope. This talk starts with observability: if you can't see what's happening in production, nothing else matters. From there, we build the eval flywheel — a practical pattern where production observability feeds error analysis, error analysis generates eval cases, and eval cases prevent recurrence.Feb 2026
"I'll just write pytest tests for my LLM"—but should you? This talk untangles benchmarks, evals, and guardrails: three concepts that sound similar but map to different Python patterns. Learn why pytest CAN work for evals (with the right mindset), why guardrails aren't tests at all, and a grounded theory approach to defining what "good" actually means for your task.
Feb 2026
Most teams approach LLM evaluation like test-driven development: write tests first, then build. But LLMs have an infinite failure surface — you can't predict what will break. This talk argues for a different approach: deploy first, observe failures, then build evals for the patterns you've actually discovered.Feb 2026
A presentation on parallel vibe coding techniques, presented at TopMonks Caffé meetup.Jan 2026
A presentation on learning with AI for high school students, presented at Zvídavá nadílka 2025.Dec 2025
A closing panel at Data Day 2025 with Kateřina Lesch and Daria Hvížďalová, organised by Jiří Vicherek, on the question "What Makes Us Human?"Nov 2025
A practical framework for introducing dbt in your org — choosing the right entry point, starting with the right features, and aligning rollout with organizational readiness.
Oct 2025
A presentation on simultaneous vibe coding, presented at AI Bootcamp 2025.Sep 2025
A presentation on evals, presented at Hacker Camp 2025.Aug 2025
A presentation on the use of Pydantic in a variety of contexts, and how to avoid its pitfalls.
Jul 2025
PydanticAI-based debate case-maker, presented at AI Tinkerers Prague.Feb 2025
A showdown of LangGraph and AutoGen, two Python-based multi-agent frameworks.Nov 2024
A presentation on the use of language models and test-driven development in software development.Apr 2024
How did I come to find myself on a stage, talking about data? How does anything ever happen, really?
Mar 2023
Prezentace pro středoškolské debatéry na téma umělé inteligence a jejích hrozeb (na začátku ledna 2023).
Jan 2023
A presentation on the use of dbt, DuckDB, and Meltano/Dagster in academic research.
Oct 2022
Coming into a company as the first data hire can be overwhelming - it certainly was for me! This lightning talk aims to distill the most important lessons for those who are considering the jump. (dbt can help.)
Sep 2022
A presentation on my personal journey from neuroscience to data science, and the lessons learned along the way.
Jul 2022
A half-day workshop at the Society for the Improvement of Psychological Science 2017 meeting.Aug 2017
A lightning talk presenting PsychTaskFramework at the Society for the Improvement of Psychological Science (SIPS) 2017 meeting.Aug 2017
¶Make your event unforgettable.Get in touch ↗