I'm a Senior Software Engineer at Bloomberg and former Microsoft, Oracle, HSBC, and Nutanix engineer. I build enterprise AI and knowledge systems: Graph RAG, agentic workflows, high-throughput data platforms, and the infrastructure that makes them dependable in production.
My favorite problems live between a promising model and a production-grade system—retrieval quality, data freshness, latency, observability, safety, and the developer experience around all of it.
| What I've been working on | |
|---|---|
| 40+ | Heterogeneous data, documentation, and metadata sources feeding Bloomberg's enterprise AI knowledge layer |
| ~500 GB/day | Data-quality and validation workloads running through 50+ Kafka consumers and Apache Spark |
| 100+ | Continuously curated knowledge bases supporting Graph RAG and agentic workflows |
| 10M+ LOC | Repository-scale AI understanding and migration work during my time at Microsoft |
| ~300 days saved | Automated C#/.NET Framework → .NET Core transformations at ~91% backend accuracy |
| 50+ repos | AI-powered code review that helped reduce PR dwell time from 12 hours to 8 hours |
I care about systems that are not only intelligent, but measurable, explainable, fast, and operationally boring.
- 🎓 BITS Pilani — B.E. Computer Science, Minor in Finance, Distinction (9.10/10).
- 🌱 Open-source contributor, currently exploring LLM tooling and systems code.
- 🎾 Away from the keyboard: tennis, squash, and badminton.
| Project | What it is | Stack |
|---|---|---|
| ♟️ ChessGrind | A focused chess-puzzle training experience with gameplay insights and a polished interactive board. | Next.js · React · TypeScript · chess.js |
| 🎨 Shader Forge | An interactive WebGL shader studio for experimenting with visual effects directly in the browser. | WebGL · Next.js · React · TypeScript |
| 🤖 LLM Experiments | A growing lab for exploring language models and practical generative-AI workflows. | Python · LLMs · ML |
| 🧩 System Design | Notes and implementations for reasoning about scalable systems and engineering interviews. | Architecture · HLD · Distributed Systems |
I contribute fixes where correctness lives in the details. Most recently, I improved Soup's MCP errors so rejected export formats and GPU choices report the original value and the live allowlist—with regression coverage for bounds, control characters, and client/server round trips.
AI & knowledge systems
Data & distributed systems
If you're working on AI infrastructure, developer tooling, or an interesting open-source problem, let's talk.
LinkedIn · Email · Repositories
Engineering with curiosity, precision, and an unreasonable dislike of flaky systems.
This is a beginner friendly repository made specifically for Hacktoberfest that helps you get your first PR.