Anastassios Nanos

Anastassios Nanos

CHIEF SCIENTIST NOFire.ai
Anastassios Nanos

Anastassios Nanos leads the Runtime team at NOFire.ai, owning urunc and the company's policy and JIT-control enforcement. His work centres on secure execution environments — microVMs, unikernels, and single-application kernels — and on bringing that isolation spectrum to ordinary OCI workflows. He holds a PhD in Computer Engineering from the National Technical University of Athens, and his research interests span systems software, virtualization, operating systems, containers, and orchestration. Before NOFire.ai he founded and built Nubificus, working on unikernel and accelerator virtualization across cloud and edge. He can be found on GitHub as @ananos.

CONFERENCE DAY 2 - Thursday 22nd October

9:30 AM Morning Keynote Presentation: Production Operations in Secure Sandboxes for Controlled Autonomy at the AI Era

Would you let an AI agent run your production operations? What happens when an autonomous agent executes an unexpected rm -rf /prod?

An AI SRE must start from what is true: a deterministic context layer, not scraped text. A time-versioned graph of Kubernetes, cloud resources, code, runtime traffic, and observability provides the foundation for pushing what's possible. Only then it can autonomously resolve an incident. 

What happens when this agent goes rogue? MicroVM sandboxing can solve part of the problem. The rest is policy enforcement outside the agent's execution environment.

Join this presentation to learn more about:

  • Why change-event causality, can move an AI SRE from describing an incident to resolving it
  • Why ephemeral microVM sandboxes can safely constrain AI operations 
  • Why policy enforcement should apply outside the agent's execution environment

Check out the incredible speaker line-up to see who will be joining Anastassios.

Download The Latest Agenda