ML Researcher and Gamer

I am a ML Reseracher primarily working in 4 different areas : 1) Vertical agents and applications, 2) Multilingual LLMs and Model Merging, 3) Agent Benchmarks and frameworks, 4) Reasoning approaches. Recently, I was the NLP Research lead at Traversaal.ai, On weekends I work on open source projects on Agents, Post-Training and Reasoning through Cohere Labs where I lead the ML-Agents group.

I have started my career through an esports org Black-Ops esports which I founded and managed for 3.5 years. We mostly earn revenue by helping connect newer content creators with advertising brands and taking a cut of their ad-revenue. I occasionaly still help them and other Instagram meme pages with content ideas and growth startegies.

Latest CV >>
Twitter >>
Huggingface >>
Linkedin >>
Github >>
Google Scholar >>

Related to e-sports

Discord >>
Instagram >>

Join ongoing projects :

Cohere Labs >>

>> As of Oct 15, Started looking for new Full time opportunities

Educational Journey

Masters in Machine Learning

University of Maryland, College Park

2022 Aug - 2024 May

Graduate Certificate in Data Science

University of Maryland, College Park

2022 Aug - 2023 May

2018 Aug - 2022 May

Bachelors in Computer Science

RVR & JC College of Engineering

Experience


Community Lead [ML-Agents]

Cohere Labs

May 2025 - Present

Leading community led open source projects on Agents, Frameworks, Reasoning and benchmarks. You can join here >>> https://discord.gg/ckaQnUakYx


NLP Research Lead

Traversaal.ai

Sep 2023 - Aug 2025

Lead a team of 15 working on Vertical agent applications, Multi-agent collaboration frameworks and tools. Multilingual LLMs for south-asian languages and benchmarks for fine-grained evaluation of Agents.


ML & IoT Intern

Bolt IoT

Feb 2021 - Apr 2021

Built Time-series models for a smart humidity, lighting, and temperature management based on current and estimated usages and periodic trends


Founder

Black Ops Esports

Nov 2018 - Apr 2022

Established and led an E-sports organization of over 200 people. Built Image-Time Series models for forecasting end results of battle royale games. Assisted gaming content creators in improving user engagement and ad-revenue.


Interested in collaborating ? Have any question regarding one of the previous papers ? you can fill the form below