hi, my name is
sharun naicker.
i like to build stuff
i'm a software engineer at JPMorgan Chase. full-stack background, with a soft spot for data and backend systems. i care about building things that are efficient, reliable, and built to last.
About me
I'm a software engineer with a background in Computer Science from the University of Texas at Dallas. I'm currently at JPMorgan Chase, working on the backend infrastructure that keeps financial data flowing reliably at scale.
I love the intersection of data engineering and product — designing the invisible systems that make applications fast, reliable, and insightful. I care deeply about writing clean, maintainable code that solves real problems.
I enjoy being active so outside of work, you'll find me rock climbing, running, or lifting. If I'm not doing those I'm probably watching anime, reviewing movies, or playing modded Elden Ring
Technologies I work with:
- ▸Java
- ▸Python
- ▸TypeScript
- ▸JavaScript
- ▸C++
- ▸SQL
- ▸MongoDB
- ▸AWS
- ▸React
Where I've worked
Software Engineer @ JPMorgan Chase
August 2025 — Present
- ▸Optimized bulk data validation for jobs processing 1K–2K records by redesigning validation checks from per-record to batch execution, reducing processing time from 150s to 67s (~2.2× faster) and enabling bulk job creation via API.
- ▸Architected migration of audit report storage from relational databases to AWS S3, reducing costs by ~30% and enabling scalable handling of frequently generated reports.
- ▸Built and integrated Java-based APIs to manage FX trade rules, connecting with downstream services via AWS-based messaging and API gateways to support event-driven transaction workflows.
- ▸Designed a Java-based event-driven API integrating Kafka for real-time data routing across enrichment, gateway, and relational database layers, reducing manual account lookup time by 60%.
Things I've built
Featured Project
PolicyGraph AI
LLM-powered agent workflow application for policy document analysis. Full-stack system with a FastAPI backend, PostgreSQL with JSONB/GIN indexes for semantic search, Redis for caching, and a React 18 frontend. Integrates with the Anthropic Claude API to power multi-step reasoning over complex policy documents.
Featured Project
UTD Building Classifier
iOS app that identifies 11 UTD campus buildings in real time using a custom-trained Core ML model. We collected 500+ on-campus photos ourselves, trained a neural network with augmentations (blur, rotation, etc.) using Create ML, and integrated S(Casp) common sense reasoning to improve classification accuracy. Built with Swift, SwiftUI, and Apple's Vision Library.
Other Noteworthy Projects
Spotify Group Controller
Web app that lets groups collaboratively control Spotify playback. Multiple users can queue tracks, and control music together in real time via a shared session.
iPlan
Mobile app for collaborative event planning with budget tracking, itinerary building, deadline calendars, and task prioritization.
04. // What's next?
Get in touch
I'm currently looking for new opportunities in the Seattle area. Whether you have a question or just want to say hi, my inbox is always open.
Say hello →