How you get to the equipment
Members and clients can use the facility by individual agreement. It is not an automatic entitlement; purpose, capacity and data handling decide.
About the Lab
The facility and the people
The Lab rests on two things: equipment you can actually measure on, and people who know what to measure and how to read the result.
The technical facility is in Ostrava. It serves model comparisons, load testing and checks of local deployment, the work that cannot be done on paper.
The exact configuration, free capacity and access conditions are confirmed for each brief. Equipment on temporary loan from partners is distinguished from the Lab's own.
01 / Equipment
The technical facility
According to the Lab's founding document the facility consists of the following. Parameters and availability are checked for each test.
- Server room in Ostrava
- Our own space with public IP connectivity, running the Lab's test and production systems.
- Server cluster
- Compute capacity for load tests, configuration comparisons and longer runs.
- NVIDIA DGX Spark
- Several units for working with models in an environment close to a local company deployment.
- Partner equipment
- Servers and other hardware on temporary loan from technology partners; availability depends on the individual loan.
- Remote access for a specific test.
- Allocated compute for an agreed period.
- Preparation and execution of a test by the Lab team.
- A loan of selected equipment or a server.
Before access is granted we define the purpose, the duration, the available capacity, the price or membership allowance, the access method and the data handling rules.
02 / Team and members
People
The Lab brings together developers, AI and infrastructure specialists and people from companies. Listed below are members whose involvement is confirmed and verifiable.
Founding member · the Lab's first member
Lukáš Orčík
AI agents · DevOps · MLOps
He has worked in IT for more than twenty years, moving from development and academic research through Big Data and cloud infrastructure to DevOps, MLOps and artificial intelligence. He has taken part in projects at VŠB-TUO and IT4Innovations, at Merck and Cognizant, and in public administration projects.
Today he combines experience running production systems with the development of AI agents and multi-agent systems, from design to deployment.
What he works on in the Lab
- Design and deployment of AI agents and multi-agent systems.
- Connecting company applications to AI through MCP connectors.
- Running and automating the infrastructure for tests and deployments.
03 / Community and taking part
Community and membership
Members bring practical problems, domain knowledge and experience. They help decide which topics are worth testing.
- Who can join
- An individual, a company or another organisation, also outside the Moravian-Silesian Region. First members join by invitation or after an individual review.
- What members bring
- Real briefs, knowledge of their field, hypotheses and test scenarios, expert feedback and a willingness to share findings within the agreed confidentiality.
- What membership brings
- Contact with people solving similar problems, the option to propose topics and experiments, and use of the facility under the agreed rules and capacity.
- What is still being prepared
- Membership tiers, their prices and exact rights are not set yet. We therefore describe the currently available options to each applicant.
Membership by itself does not define the scope of a paid service. Tests, development and reserved compute are agreed separately.
04 / Services
Other services of the Lab.
The services build on each other. You can start with any of them, depending on where you are.
Validating an AI idea
Can AI handle a task our people do today?
Service detailsComparing models and hardware
Which solution has the quality and handles our workload?
Service detailsDeployment design and prototype
Should it run on-premise, in the cloud, or both?
Service details