Mirek Umlauf
Fractional Chief Data & AI Officer · Board Advisor
Data and AI in the places where technology and people both decide.
Twenty years delivering data and AI in global, complex environments — two IPOs (NYSE, then London), dozens of integrated acquisitions globally, Chief Data Officer of a global consumer cybersecurity business. The industries varied; the pattern didn't.
Today I sit on the other side of the table — alongside boards and leadership teams, often in companies that didn't grow up with large data teams. Teaching and research at FIS VŠE keeps the practice honest.
Count on me to ask the hard questions before the expensive decisions: name the real problem, size the technology to it, match the architecture to your skills — and choose deliberately what gets locked in. Commitment can build advantage; it just shouldn't be a vendor choosing where you commit.
Geek – Nerd – Suit: fluent in research, fluent in data, fluent in the boardroom.
Companies where data and AI are decided by people who understand both the technology and the business — and where working with them makes people grow.
Execute at a cost the business would choose again. Bring what research actually knows into the Monday-morning business. Leave every team — and myself — more capable each year.
Expertise
Research in one hand, the P&L in the other
The gap in most companies isn't awareness of AI — it's practical use. That's the gap I close, and not only for AI: for every data technology the business already owns. I apply research to businesses — bringing what the science actually knows into functions that have to use it Monday morning — and I develop the same people to get measurably better in both halves of the job: the technology and the leadership.
Patterns — from architecture to the anatomy of the firm
Technology has proven architecture patterns; so does everything around it — how teams are shaped, how change takes hold, how business model and process connect to data. I diagnose them together, as one anatomy. Structure creates outcomes: everyone says they want to be data-driven; few build the structure that produces it. That's why the fix is usually not another platform.
Designing the data you'll wish you had
A "data problem" is usually a problem about how the company sees itself — what counts as a customer, where the margin sits. I know what to ask the data, which answers to mistrust — and that missing data can't be backfilled. So I design collection and capture — process events, customer signals, IoT and sensor data — for two readers: your people, and the AI agents that help them run the business. Both need to know, fast, what's actually going on.
AI and decision systems for operations
Decision-intelligence layers on production, operations, and customer data — systems that help people see earlier and decide faster. Especially in manufacturing and industrial environments, where data and decisions are most directly tied.
Data & AI governance, non-invasive
Data treated as a product, ownership inside operations, governance that formalises the accountability already in the business. The discipline I also serve on the board of DAMA Czech Republic.
Teams that survive me — and experts I don't have to search for
Engineers, analysts, and platform people who deliver in operationally-led environments. When the work needs a true specialist, I bring in someone I've known for years — not a CV from a database. And I work through the incumbent CIO and IT leadership, not around them.
Work & ventures
- Construction logistics group, €1bn — IT and operating-model transformation, Conway's law in practice
- Global industrial controls manufacturer — introduction of BI
- CEE pharmacy retail chain — data governance and big data training
- Private clinic — multi-year ~$10M digital transformation
- Mid-sized manufacturer — IT transformation
- Manufacturing companies — customized data modeling training
- Marketing research firm — data warehouse consolidation
- European fintech — first analytical data layer
- Legal tech — knowledge engineering: ontologies and knowledge graphs for complex legal agendas
Selected outcomes
- A data warehouse the client was about to buy — and didn't need. Avoided: ≈ €200,000.
- A transformation programme reduced by 40% — with no loss of outcome.
- Millions of USD in licence savings — after migrating a large-scale data warehouse to open source.
- Hundreds of spreadsheets retired — right information, right time, one source.
- Month-end close from day 15 to day 3.
- A full data warehouse consolidating two ERPs, first executive dashboard in production — three months from contract.
- IPO data readiness in four months — passed every audit.
- Data due diligence for an acquisition — delivered in three weeks.
- Patient analytics for 3,000 patients a month — healthcare-grade privacy compliance at €3,000 a year.
- An internal team — business and IT trained together — runs their data warehouse themselves.
Named references — and my exact role in each outcome — available in conversation.
Entrepreneurship is the other half. I've co-founded four data companies. One failed and paid for the lesson; Nuefis works and stays profitable; two are in build — legal tech and board tech. No unicorn stories.
Teaching & community
FIS VŠE — Assistant Professor
Leadership in IT, Enterprise and Data Architecture, Lean Manufacturing and Data, Data-driven Decision-Making in Manufacturing — across the master's and IT Management MBA. Research projects and bringing AI into teaching itself, working with large automotive groups, companies like Siemens, and vendors of data technologies.
Controller Institut — hard data-skills training and data analytics
Years of training practice before academia, under the Controller Institut (controlling.cz). Together, 1,500+ executives and students.
Mentoring
Board and management members, FIS students, career inflection points via Femme Palette. My own training: IMD and UC Berkeley.
DAMA Czech Republic — Board Director, Sponsorships & Events · volunteer
Advancing data & AI governance as a discipline in the region.
IMD Alumni Club CZ & SK — President, volunteer
A network of people who share the belief that learning doesn't stop at the diploma.
A few clients at a time, by design — and everything documented, so nothing depends on my calendar.
No newsletter, no funnel. If you're facing a decision like these, write me: mirek@umlauf.tech