About Me

Hi, my name is

Videet Mehta.

I'm a student at MIT studying Computer Science. I'm passionate about frontier AI research in multi-modal LLMs and hardware/software model acceleration.

I'm currently working at HAN Lab at MIT, where we are attempting to accelerate diffusion language models. I'm also doing research at MIT's Spoken Language Systems Lab under Jehanzeb Mirza on finding optimal attention heads for audio event classification & spoofing detection.

Previously, I worked at Mercuria Energy trading where I worked on forecasting marginal prices. Additionally, I did some work at Sarvam AI to build India's first conversational speech AI in Hindi and English.

I'm also proud to have previously represented USA in the International Olympiad in Artificial Intelligence in 2024 and to have won a gold medal! I am now on the organizing and scientific committee for the 2026 USA AI Olympiad Team.

Here are a few technologies I've been working with recently:

  • PyTorch
  • JAX
  • DDP
  • Deepspeed
  • CUDA
  • SQL
  • Node.JS
  • React
Check out my Github!

Work Experience

  • AI Researcher @ H.A.N Lab
    Sep. 2025 - Present
    MIT

    Working with Qinghao Hu on accelerating diffusion language models to match the accuracy of similarly-sized auto-regressive models before we tackle speed. Efficient AI!

  • Machine Learning Engineer @ Sarvam AI
    June 2025 - September 2025

    Building Hindi and English speech-to-speech foundation models and large-scale distributed training infrastructure.

  • AI Research Scientist Intern @ Mercuria Energy Trading
    March 2025 - August 2025
    Houston, TX

    Built ML systems for power markets: forecasted data-center load, optimized GPU/CUDA pipelines to halve FM fine-tuning costs, and designed graph+sequence models that improved LMP MAE/RMSE by ~25%.

  • AI Researcher @ MIT CSAIL Spoken Language Systems
    September 2024 - Present
    Cambridge, MA

    Built a PEFT multimodal AVSR pipeline with distributed training that reduced WER by 15% and prototyped sparse attention-routing on Qwen2.5-Omni for audio tasks.

  • Founding Engineer @ Hidden Studios
    Feb. 2025 - June 2025
    Cambridge, MA

    Built a full-stack in-game advertising platform with analytics, ML-based impression prediction, and automated gameplay data collection.

  • AI Researcher @ Regenerative Neurotechnology Lab
    March 2024 - August 2024
    UT Health System

    Built a Viterbi decoding pipeline with statistical LMs for neural-to-word decoding and aligned GPT-2 hidden states with EEG features.

  • AI Research Scientist @ Houston Learning Algorithms
    March 2023 - April 2024
    University of Houston

    Built a conditional video-diffusion model for real-time wildfire prediction and co-authored an IEEE MMSP 2024 paper with an arXiv preprint.

Projects

Recent Posts

View All Posts →

What’s Next?

Get In Touch

I'm always open to new opportunities and connecting with other students and professionals. Whether you have a question or just want to say hi, feel free to reach out!