~/about

aionprem.pl is technical notes on deploying AI outside the public cloud. Written for security architects, IT directors, CISOs, and compliance leads who face one concrete question: how to set up AI in an organization where data cannot go to OpenAI, Anthropic, or Microsoft, and which has to document this for an audit.

We do not sell software. We do not promise that "AI will solve compliance". We document concrete architectures, regulatory mappings, risk analyses, GPU sizing benchmarks, and reflections from deployments we participate in or observe.

Who this is for

aionprem.pl readers are typically:

• CISOs of NIS2 essential entities forming an opinion on vendor risk
• CIOs or infrastructure leads deploying AI on owned hardware without a dedicated consulting budget
• Security architects or compliance officers preparing vendor evaluations for an RFP
• Lawyers specializing in cybersecurity, GDPR, or the AI Act looking for reference materials

What's not here

• Product marketing
• Political debates about NIS2 or the AI Act
• AI for the public sector or defense (off-thesis)
• Sectors other than manufacturing, energy, supply chain, and adjacent NIS2 essential entities

Author

Fryderyk Pryjma writes about deploying AI outside the public cloud from the perspective of someone building, not selling.