Hi, I'm Denis

Platform engineer working at the intersection of production infrastructure and applied AI.

I build agent systems that reason over knowledge graphs, and the infrastructure that runs them.

What I do

• Platform engineering: AWS, Kubernetes, Terraform, internal developer platforms, CI/CD
• Applied AI: agent systems, GraphRAG, knowledge graphs, LLM observability

A support-ticket triage agent that answers one question support teams ask dozens of times a day: who should handle this?

Instead of matching keywords, Korca models a team's expertise as a Neo4j knowledge graph: who's resolved this kind of ticket, who knows this client, who owns the relevant SOP. It routes on that. In production it routes correctly more than 90% of the time.

• 📖 How it works — published on the Neo4j blog
• 💻 Source — open source, MIT
• 🥉 3rd of 40+ projects, Neo4j Aura Agent Hackathon

Stack: Neo4j Aura · Aura Agent · FastAPI · Celery · Gemini embeddings · React

A real Kubernetes cluster at home, where I run and break things before they reach production at work.

• Kubernetes with Cilium (eBPF networking), Longhorn storage
• GitOps end to end via ArgoCD
• Infisical for secrets, LiteLLM proxy + Langfuse for LLM routing and observability
• Self-hosted AI tooling, agent workloads, and whatever I'm currently testing

• Cloud & Platform: AWS (ECS, Fargate, Lambda, RDS, SQS) Kubernetes (EKS, Helm, Karpenter) Terraform/OpenTofu Docker
• GitOps & CI/CD: ArgoCD Argo Workflows GitHub Actions Infisical
• Observability: Prometheus Grafana Loki Langfuse (LLM tracing & cost)
• AI / LLM Systems: Neo4j & knowledge graphs GraphRAG & hybrid retrieval agents (PydanticAI, LangGraph) LiteLLM
• Languages & Data: Python (async, FastAPI) TypeScript/React Go PostgreSQL Redis Cypher

AWS Certified Solutions Architect

HashiCorp Certified: Terraform Associate

Neo4j Certified Professional