I design production-grade AI-native infrastructure — from self-healing Kubernetes platforms to agentic RAG pipelines and multi-model orchestration systems.
I am a senior AI/Cloud engineer specializing in building autonomous, production-grade systems at the intersection of cloud infrastructure and artificial intelligence.
My work focuses on multi-agent architectures, confidence-gated automation, and agentic retrieval systems — engineering software that can reason, decide, and act with minimal human intervention.
Every project I build is production-first: Terraform-managed infrastructure, proper observability, security boundaries, and documented runbooks — not proof-of-concepts, but deployable systems.
A production-grade AIOps platform for Kubernetes that intercepts infrastructure anomalies, routes them through a multi-agent reasoning pipeline, and takes autonomous action calibrated to confidence. Every remediation becomes a Git commit — the cluster never mutates outside a reviewed PR or a high-confidence auto-apply.
A production-grade agentic retrieval system for medical knowledge, powered by LangGraph workflows, pgvector semantic search, and streaming SSE responses. The system routes queries through a confidence-scored pipeline — iterating on relevance up to 3 times before falling back to live web search.
A unified orchestration platform for running workloads across multiple large language model providers. Designed to abstract provider-specific APIs behind a common interface — enabling intelligent model routing, cost tracking, and fallback strategies across OpenAI, Anthropic, and open-source model backends.