I design, develop, and deploy cloud-native distributed systems that solve real problems at scale. My engineering philosophy is straightforward: build systems that are simple, resilient, and fiercely cost-effective. Shipping features is only one aspect of the job; engineering high-quality, uncompromising solutions that actively drive down compute and infrastructure costs is the real metric of success.
In practice that has meant a ~4TB zero-downtime migration that cut cloud cost 2X, Kafka disaster recovery across 45+ services, and multi-tenant frameworks serving millions of users. The full track record is on Work With Me.
One production failure mode at a time, mechanism first. A new log every two days.
Why N+1 queries are invisible to every database side metric you own, how a getter call becomes a network round trip, and why the fix that looks obvious produces a cartesian product.
Why the ANSI isolation table does not describe your database, what a phantom read actually is at the index level, and why snapshot isolation still lets two correct transactions corrupt each other.
LFU evicts the least popular key, not the oldest. The O(1) version needs two hash maps and linked lists per frequency bucket.
Architecting reliable, highly-available backend systems for enterprise e-commerce. Focused on practical performance, seamless scale, and optimizing cloud infrastructure to keep operational costs strictly optimized without sacrificing quality.
Developing agent-based systems, Model Context Protocol (MCP) servers, and agent-to-agent workflows. Building tools that bridge developer, support, and operations needs by offloading routine tasks through deterministic AI orchestration.
Walmart Global Tech India
Leading agentic AI infrastructure for enterprise e-commerce: LLM pipelines and MCP server tooling in production, enterprise-scale retrieval (hybrid search and RAG), and token cost engineering for autonomous agent workloads.
Led multi-tenant framework architecture enabling new market launches at millions-of-users scale; built LLM agents and MCP server tooling automating developer, support, and data operations.
Migrated a ~4TB customer-facing Cosmos DB container with zero downtime, cutting cloud cost 2X and raising scalability 10X; planned and executed Kafka disaster recovery across 45+ services.