I design and operate
production-grade systems.
Architecture · Infrastructure · AI Systems · Platform Engineering
No boilerplate. No agency. Complete systems, from design to operation.
RUNTIME
AVAILABILITY
DOMAIN
Arch / AI / Infra
UPTIME
17+ yrs prod
Active modules
MOD_01
LOADEDArchitecture Engine
Distributed systems design, micro-services, robust APIs. From sketch to production architecture schema.
MOD_02
LOADEDDevOps Control Plane
CI/CD pipelines, containerization, Kubernetes orchestration, GitOps. Reproducible and observable infrastructure.
MOD_03
LOADEDAI Runtime Layer
LLM orchestration, RAG pipelines, explainable AI agents. Local, sovereign, controllable AI — no black box.
MOD_04
LOADEDProduct Engineering Layer
From product to run. Technical leadership, hiring, roadmap, CTO management. Systems that operate without their creator.
Deployed systems
Legibus
Architect / AI IntegratorLegal AI assistant — LLM + RAG pipeline
Metalearn
CTO / Infrastructure OperatorSovereign EdTech ecosystem on Kubernetes
MON-GN
Architect / OperatorProduction e-commerce platform — real load
Execution log
Engineering principles
RULE_01
Graceful degradation
Systems must operate partially rather than collapse entirely.
RULE_02
Complexity = liability
Every unjustified abstraction layer is a future cost.
RULE_03
Production defines architecture
Real constraints override theoretical best practices.
RULE_04
Automation & cognition
Automation must reduce cognitive load, not displace it.
RULE_05
Reliability > novelty
A boring system that works beats a brilliant but unstable one.
RULE_06
Data under control
Data sovereignty: no forced dependency, no opaque lock-in.
RULE_07
Maintainability > speed
A system shipped fast but incomprehensible is a timed bomb.
System integrations
System documentation
Architecture decisions, design philosophy, technical notes.