Sr Software Engineer · Apple  //  Full-stack · Generative AI  //  Austin, TX

AbhinayDutta.

I build AI products that feel like a superpower.

Full-stack engineer, 15 years deep in distributed systems — now building with generative AI: RAG pipelines, locally-run LLMs, and agents that take on the work nobody wants to.

LLM applications RAG AI agents MCP Full-stack Distributed systems
01

About

I understand systems deep enough to build my own when the tools fall short — and make them hold up at scale.

I've spent 15 years building distributed systems across finance, payments, and consumer tech. Today my focus is generative AI — designing and shipping AI into user-facing products, and providing technical leadership across teams that bridges AI research with what actually ships.

I value thoughtful design, developer empowerment, and building products that make people feel like they're working with a superpower. And I love the problems no one else wants to touch — from low-level email-protocol debugging to designing an agent that defuses a critical production incident.

I also treat the non-functional requirements as first-class — often above the feature itself. I design for availability, resiliency, and zero-downtime, build in automated backups and self-healing, hold a hard line on P1s, and sweat the user experience. Software earns customer trust by staying up and behaving — not just by what it ships.

LLM applications
RAG pipelines and locally-deployed models — DeepSeek, Mistral, OpenHermes via Ollama.
AI automation
Agents that take on effort-intensive tasks by leveraging MCP servers.
Developer productivity
AI woven into development workflows through IDE plugins.
Scalable infrastructure
Leading distributed architectural decisions across products.
02

Selected work

/ 01

A graph database,
rebuilt from scratch

Access control ran through RedisGraph across multiple permission layers — and it couldn't keep up. I designed and built a custom in-memory graph database to replace it, CRDT-replicated across regions for conflict-free consistency. The performance the system actually needed, at Apple scale — a from-scratch answer when the off-the-shelf tool fell short.

/ 02

Agentic engineering,
at org scale

Managed cloud agents that follow an agentic-engineering pipeline to onboard every project and repository across the org onto Kubernetes. AI agents doing the engineering work themselves — a migration at a scale that simply isn't feasible by hand.

/ 03

Real-estate AI
maintenance platform

A founder project: a LangGraph multi-agent system that triages, diagnoses, and resolves property-maintenance issues end to end — coordinating across tenants, vendors, and owners. Built for the failure modes and reliability bar that real production traffic demands.

03

Experience

AppleSr Software EngineerEmail, calendar, documentation and bug-tracking systems; large-scale migrations and performance engineering.
2020 — PresentAustin, TX
VisaStaff Software Engineer · Sr. Software DeveloperBuilt large-scale distributed systems processing payments for ~25% of Visa.
2017 — 2020Austin, TX
Bank of AmericaTechnology Associate
2016 — 2017Charlotte, NC
Texas A&M UniversityWebMaster
2014 — 2016College Station, TX
Fidelity InvestmentsAnalyst — Marketing Analytics
2013 — 2014Bangalore, IN
SapientNitroTechnology & Digital Analytics
2010 — 2013India
04

Education

Texas A&M UniversityM.S., Management Information Systems · GPA 4.0
2014 — 2016
NIT JalandharB.Tech, Computer Science
2006 — 2010
05

Get in touch

abhinay.nitj@gmail.com

The fastest way to reach me. Say hello, or bring me a hard problem worth solving.