About
I'm a Software Engineer II (MTS II) at Cohesity. I work on Gaia, Cohesity's generative AI assistant that uses RAG and LLMs to give insights from enterprise backup data. My part of it is secure, private data integration for customers and letting people ask questions about their data in plain English.
Before Cohesity, I was an Associate Software Engineer at Veritas for about two and a half years. One of the main things I built there was Alta Copilot, a multi-agent system on LangGraph with RAG, NL-to-SQL and decision-making agents. I implemented the LangGraph framework for it from scratch and also set up the LLMOps evaluation pipelines and observability. Apart from that, I worked on NetBackup and Alta View modules with Java and Spring Boot.
I studied Computer Science at Pimpri Chinchwad College of Engineering (PCCoE), Pune, and graduated in 2022 with Honors in AI and Machine Learning. College is where I got into ML. I built a Conditional GAN that colors anime sketches, a Discord bot that catches toxic messages, and wrote about GANs and ML on Towards Data Science. I was also a Microsoft Learn Student Ambassador, Management Head at Google DSC PCCoE, Technical Head and Webmaster at the PCCoE ACM Student Chapter, and a technical author for ACM XRDS magazine.
I still go back to PCCoE sometimes to take sessions for students. You can find the material from those sessions at /pccoe.
Selected talks
GAN and LLM architecture, plus a code walkthrough of a Conditional GAN, RAG, and an agentic framework built on LLMs.
ATAL Academy Faculty Development Program, Computer Vision
GAN architecture and types, with a code walkthrough of the Conditional GAN behind Sketch to Color.
Why businesses containerize, Azure fundamentals, and an intro to the Microsoft Learn Student Ambassadors program.
- End to End AI Discord BotMar 2021
Building a Text Toxicity Detection Discord bot with discord.js and deploying it to Azure App Services.
- Pie & AI: Pune - Intro to GANsOct 2020
DeepLearning.AI meetup, hosted by PCCoE ACM Student Chapter
A beginner-friendly walkthrough of GANs, their types, and where to go from here.
- Opensource - Why and How?Sep 2020
DSC PCCoE, with PCCoE ACM Student Chapter
A hands-on workshop for getting started with open-source collaboration.