I build multi-tenant, event-driven backend platforms for identity and access management — systems that stay correct when things go wrong. 8+ years of shipping production software in Java and the cloud.
I design backend systems that have to stay correct when things go wrong: requests that span many services, many tenants, and many approval steps, without losing track of a single one.
Most of my recent work sits where identity security meets platform engineering, at Saviynt in Los Angeles. I like owning a system end to end — from the architecture review to the production fix — and helping the people around me grow through design reviews, code reviews, and mentoring.
I hold an M.S. in Computer Science from Cal State East Bay. Outside work, I judge hackathons and enjoy talking with other engineers about distributed systems.
Kafka and RabbitMQ pipelines with built-in recovery, so no request gets stuck.
Access requests, approvals, provisioning, and identity proofing at enterprise scale.
Isolated data per customer, cloud-native on Kubernetes and AWS.
Backend engineer on Saviynt's identity security platform, now on the Edge Platform team.
Architected and built the service that takes an access request from submission through multi-level approval to provisioning, with an isolated database per customer. It covers applications, roles, entitlements, AI agents, and MCP tools.
Designed self-recovering request flows that removed stuck-request incidents, and cut auth load on the database by ~99%.
Led the backend for a team of ten onboarding 150K+ external identities. Moved slow synchronous flows to async processing, added a pluggable identity-proofing framework with two vendors, and led the move to a modern workflow engine with a no-code form builder.
Telecom product catalog with thousands of 5G, broadband, and WiFi offers, used by AT&T and two other major operators. Team of seven engineers.
Mediamorph, a content value management platform used by media studios including Disney, Warner Bros, and FHE.
TCS MasterCraft DataPlus, an enterprise data-quality product.
Judge
Judged AI projects on demo day. Standouts included smart glasses with on-device help, instant 3D room redesign, cheap LoRA specialists in place of frontier-model calls, and drone hazard detection streamed to smart glasses.
The best teams engineered around cost, latency, and real constraints instead of chasing the newest model.
Judge
Represented Saviynt as a judge for 300+ projects built in 36 hours, alongside judges from Uber, Google, and OpenAI. Standouts included Cropt, my personal favorite Snitch for the real-world problem it solved, and CodeBreaker, the overall winner.
The best engineering starts with finding a problem worth solving, then having the courage to build it.
Hacker
Built a healthcare-access app with a team of four, helping people in underserved communities find affordable medication nearby for everyday issues like headaches and body pain.
Designed it to send anything serious straight to 911 or the nearest emergency room.
Attendee
Explored talks and demos on treating APIs as products, AI-generated database code with Couchbase, Fluree's knowledge graph database, and StackHawk's API security testing in CI/CD. StackHawk featured me on their LinkedIn page.
Biggest takeaway: treat an API like a product. Define a true MVP, then iterate.
California State University, East Bay
2022 – 2024 · GPA 3.8
Pune Institute of Computer Technology
2013 – 2016