Gateways, runtime planes, MCP governance, identity, quotas, and how private AI stays operable on Kubernetes.
Available for conversations
Meet Andrey Lesnikov
Talk through architecture, AI infrastructure, and platform reliability — calm systems, clear trade-offs, production evidence. Based in Halle (Saale), Germany; remote conversations welcome.
01
Good fit
What these conversations are for.
Short, concrete working sessions — not a sales pitch. Bring a system boundary, an incident pattern, or a platform decision you need to pressure-test.
EKS foundations, Argo CD / Flux ownership, promotion gates, drift, rollback contracts, and delivery evidence.
Restore automation, rehearsal before deploy, observability meshes, SLOs, and incident paths that stay calm under pressure.
02
Format
How a session usually runs.
- Length
- 30–45 minutes for a focused architecture or platform discussion.
- Preparation
- A short problem statement, current constraints, and what “good” looks like for you.
- Outcome
- Clear trade-offs, a preferred path, and links to relevant public cases or repositories when useful.
- Languages
- English or Russian.
03
FAQ
Before you book.
No. It is a working conversation about infrastructure boundaries, trade-offs, and evidence — not a product pitch.
A short problem statement, current constraints, and what “good” looks like. Links to architecture notes help.
Usually via LinkedIn, GitHub, Telegram, X, or Xing — or a follow-up call after first contact through the Meet channels on this page.
Available for meaningful infrastructure conversations