name : Thakur Prasad Rout
role : Platform & Site Reliability Engineer
company : iServeU | Bhubaneswar, Odisha
education : B.Tech — Electronics & Telecommunication, VSSUT (2024)
currently_building:
- 3-zone multi-master Kubernetes HA cluster (STPI DC + On-Prem + GCP)
- Cilium CNI (eBPF) + Cluster Mesh — zero-trust production networking
- RAG AI chatbot in Rust (Weaviate + on-premise Ollama + MongoDB)
- PodXS — browser-accessible VNC desktop platform on K8s StatefulSets
passionate_about:
- eBPF & kernel-level networking (Cilium, DeepFlow)
- SLO/SLI-driven reliability & on-call toil reduction
- High-performance systems in Go & Rust
- AI/ML infrastructure — RAG pipelines, on-premise LLMs
hobbies : [Dance, Exploring Colors, Breaking prod (and fixing it faster ⚡)]|
Production-grade multi-master K8s across STPI DC, on-prem & GCP with Cilium CNI (eBPF), Cluster Mesh & zone-aware scheduling — zero data-plane downtime on full zone failure.
|
High-perf RAG backend in Rust (Axum + Tokio) + Weaviate vector search + Ollama on-prem LLM. 75% less memory · 10× faster cache · 50× faster startup vs Node.js.
|
|
Browser-accessible isolated dev desktops on K8s StatefulSets (KasmVNC + XFCE4). GitOps ConfigMap lifecycle, clipboard exfiltration controls & Node.js/React admin portal.
|
Replaced Apigee in production with WASM-filtered Envoy + Cilium Gateway API — ~40% cost reduction while routing 100% of production traffic.
|
Languages
Kubernetes · Cloud · Infrastructure
Observability · CI/CD · DevOps
Databases · Tools