ABOUT

I build AI systemsthat have to workin the real world.

I'm a Tech Lead and AI Engineer working at the intersection of AI systems, distributed architecture, and product engineering.

Over the last 9+ years, I've moved from mobile and full-stack development into building AI-first products and production-grade backend systems.

Snehal Agrawal

Building AI products, not just prototypes.

My work today focuses on taking AI systems from an idea to something that can actually operate reliably in production.

That means thinking beyond the model itself — product requirements, architecture, data flows, agent orchestration, reliability, performance, infrastructure, and the engineering practices required to keep everything running.

I currently work on Hupo's AI sales-enablement platform, building AI roleplay and coaching experiences for enterprise sales teams across Asia.

Areas I keep coming back to.

01

AI Systems

LLM-powered applications, RAG, structured outputs, evaluation, and systems built around models.

02

AI Agents

Agent orchestration, tool use, multi-step workflows, memory, and making agentic systems predictable enough for production.

03

Distributed Systems

Asynchronous processing, event-driven architecture, queues, workers, streaming systems, and high-throughput backends.

04

Production AI

Reliability, latency, observability, infrastructure, scaling, cost, and the engineering details that demos usually hide.

From full-stack engineering to AI systems.

2017 →

Full-Stack & Product Engineering

Started with application development across mobile, frontend, and backend systems, building a foundation in product engineering and software architecture.

2024 →

Real-Time AI

Led engineering for real-time AI voice systems combining speech recognition, LLMs, text-to-speech, streaming infrastructure, and Kubernetes.

2025 →

Agentic AI & Distributed Systems

Worked on scalable agent platforms, multi-step LLM workflows, asynchronous execution, agent versioning, and AI-powered systems designed for production.

Today

AI Product Engineering

Building AI products end-to-end — from shaping product requirements to backend, frontend, architecture, and the engineering practices required for enterprise production environments.

ENGINEERING PHILOSOPHY

The interesting part of AI isn't just making a model respond.It's building everything around it that makes the system useful.

That's what I enjoy working on — the layer between an impressive demo and a system people can actually depend on.