AtharvLingayat

AI Engineer@ ContentstackMumbai, India

I engineer agentic systems that go beyond LLM calls, with tool orchestration, stateful workflows, and evaluation-driven reliability. From system design to execution, I build AI that works in production.

View selected work→
Get in touch
Portrait of Atharv Lingayat
Agentic AIRAGTypeScriptReactNext.jsPostgreSQLGraphQLNode.jsDesign SystemsAgentic AIRAGTypeScriptReactNext.jsPostgreSQLGraphQLNode.jsDesign Systems

About

The interesting problems live beneath the surface.

I'm a software engineer focused on system design, backend architecture, and agentic AI, building and engineering robust systems that turn complex ideas into reliable, production-ready software.

Experience

  • Associate Software EngineerAug 2025 - Now

    Contentstack

  • Software Engineer InternJan - Jul 2025

    Contentstack

  • Web Developer2023 - 2024

    Trikuta Infotech

Now exploring · Agentic UI & real-time generative interfaces

Selected work

Things I've built.

Products, tools, and experiments I've designed and shipped end to end.

See all projects→
01AI Assistant · Platform
↗
Polaris chat assistant running inside the Contentstack CMS

Polaris

Contentstack's context-aware AI assistant. I build its Mastra agents, from tools over product APIs to the evals that gate them, and architected its AG-UI frontend.

Mastra · LLM agents · Node.js · TypeScript · AG-UI · React · rspack
02Applied AI · Micro-frontend
↗
Ask AI generating a content model inside Contentstack

Ask AI

An embedded micro-frontend that brings conversational, AI-assisted content modeling into the host app, streaming generated schemas into an interactive, editable tree in real time.

React · TypeScript · Zustand · Micro-frontend · Playwright
03Open Source · Developer Tooling
↗

LLM JSON Validator

A zero-dependency TypeScript library that repairs and completes incomplete JSON streamed from LLMs, quote balancing, bracket healing, and append-aware parsing.

TypeScript · Streaming · npm · Zero-dep
04Developer Tooling · Side Project
↗
SSE Formatter turning a raw Server-Sent Events stream into a readable timeline

SSE Formatter

A developer tool that turns a raw Server-Sent Events stream into a readable, searchable timeline. Built for debugging agent and LLM runs where events otherwise fly by as an unreadable wall of text.

Next.js · React · TypeScript · Tailwind CSS

From the journal

Thoughts, in ink.

All writing→
Streaming JSON From LLMs Kept Breaking My UI
AI Engineering01
6 min

Streaming JSON From LLMs Kept Breaking My UI

npm i llm-json-validator. Not a typo. Here's the bug behind partial JSON from LLMs, why it wrecks your UI, and the little buffer I built to fix it.

Read article→
Micro Frontends: The Magic of Decoupling
Architecture02
6 min

Micro Frontends: The Magic of Decoupling

Breaking a monolithic frontend into independently deployable apps, how we wire them up with Module Federation, and how the pieces talk without stepping on each other.

Read article→

Let's collaborate

Have a project
worth obsessing over?

Whether it's an agent that needs a real interface, a streaming UI that keeps breaking, or an idea worth prototyping, I'd love to hear about it.

Start a conversation→
Emailatharvofficial7@gmail.com↗GitHub@atharv2608↗LinkedInin/atharv-lingayat↗
Local time--:--:-- IST
Usually replies within a day