01 Full Stack AI Engineer

Buildingcapablesoftware.

I work across product, systems, and intelligence to make software useful in the real conditions it has to survive.

Based inChennai, India
Current focusAI-native product engineering
Open toHigh-growth engineering work

A short introduction

I am drawn to work that is larger than a feature: products with real users, meaningful data, and decisions that have to hold up.

My route began with automation and IoT, moved through enterprise product engineering, and is now focused on agentic AI.Early in my career, I deliberately choose projects that expand what I can build, test, and understand.

02 / Experience

Built inside products that matter.

Hands-on engineering across AI products, SaaS, enterprise platforms, research, and automation.

2025 - Now

Founding Engineer Instalant · Sacore AB

Building an AI-powered sourcing platform end to end - campaigns, workflows, inboxes, integrations, and the agent layer behind them. The work includes MCP, Mastra, structured tool contracts, evaluation, and a faster multi-provider enrichment pipeline.

  • Next.js
  • MCP
  • Mastra
  • PostgreSQL
  • AWS

2024 - 2025

Software Development Engineer II Gooseberry Technovision

Shipped data products, a refinance micro-SaaS, Salesforce-connected automation, and private AI workflows - working where product delivery, integrations, and reliability meet.

  • Angular
  • Next.js
  • n8n
  • Ollama
  • WebSockets

2022 - 2024

Software Analyst TVS Next

Worked on healthcare and HRMS platforms, including frontend architecture, analytics, secure payload handling, and an IoT inventory proof of concept that reduced delivery time by 40%.

  • Angular
  • Node.js
  • D3.js
  • Three.js
  • Azure DevOps

2021

Research Intern IIT Indore

Developed a semi-automated textile pin-arrangement system through simulation and virtual prototyping, targeting roughly 12,000 pins per hour per machine.

  • Automation
  • Simulation
  • Virtual prototyping

03 / Projects & research

Ideas made concrete.

Public work, research, and challenge builds that have shaped the way I approach complex systems.

01 / Product

Cognaxis

An AI-assisted journal exploring permission-scoped memory, trustworthy context, and the boundaries of a system that remembers.

Gemini · Firebase · Cloud Run · TypeScriptView project ↗
02 / Challenge

Unmanned rover concept

Led a six-person team at the IET-NASA Space Apps Challenge. The concept earned the People’s Choice Award and third prize.

Systems thinking · Team leadership · Space Apps 2020
03 / Research

Automatic flushing system

Hydraulics-based research presented and published through ICTAME and IRJET in 2021.

Research · Hydraulics · Publication
04 / Prototype

IoT inventory system

A proof of concept using connected devices and a product interface to reduce floor footprint, manual effort, and delivery time.

Ionic · React · Elasticsearch · Arduino

04 / Technology

Tools I reach for.

A working set that changes with the problem, not a list for its own sake.

Agentic AI

MCP, Mastra, LLM and SLM applications, orchestration, tool calling, context, memory, evals, and human-in-the-loop patterns.

Product platform

TypeScript, JavaScript, Next.js, React, Angular, Hono, NestJS, Express, APIs, microservices, and real-time systems.

Experience layer

Tailwind, Three.js, D3.js, WXT, browser extensions, Playwright, Puppeteer, and thoughtful interfaces around complex workflows.

Systems & delivery

PostgreSQL, MongoDB, Redis, AWS, Docker, CI/CD, OAuth, JWT, RBAC, TDD, and automated end-to-end validation.

05 / Recognition

Milestones worth keeping.

Programs, awards, and research that sit alongside the day-to-day engineering work.

2026

Google GenAI Academy

GenAI Master. Cognaxis was developed through the program.

2020 - 2021

IIT Madras NPTEL

Operations Management specialization, including Top 1% and Top 10% results.

2020

IET-NASA Space Apps

People’s Choice Award and third prize for the unmanned rover concept.

The next thing should be harder.

I want to contribute to AI products with meaningful data, consequential workflows, and enough ambition to keep the learning curve steep.

Start a conversation