Who Am I?
Hi! I'm Sathya,
I'm an Engineer who loves exploring how computers can understand and work with
different types of information - whether it's text, images, or speech.
I enjoy solving challenging problems and collaborating with people who share my curiosity.
To be a bit more technical, I like working on multimodal multilingual machine learning models,
evaluating, and improving their performance.
Originally from Chennai, India and currently based in Mountain View, USA I spend most of my days working at Google. Previously I was at Carnegie Mellon University working on projects in artificial intelligence.
When I'm not working on projects, you'll probably find me on the cricket field or playing badminton. I'm also a huge movie buff and pride myself on enjoying both critically-acclaimed masterpieces and what others might call "terrible" movies - because sometimes those are the most fun! I believe there's charm in both a well-crafted film and a so-bad-it's-good movie marathon.
Education
Carnegie Mellon University
- Machine Learning track
- Relevant Topics: Machine Learning, Deep Learning, Natural Language Processing, Multimodal Machine Learning, Multilingual Machine Learning
Work Experience
Machine Learning Engineer
JPMorgan & Chase
Jan 2025 - Mar 2025- A very small gig at JPMC working on personalizing the Chase Banking Application
Research Associate
Carnegie Mellon University - Language Technologies Insititute
Jun 2024 - Dec 2024- Worked on Multimodal Multilingual problems with Prof. Graham Neubig
- Researched on image transcreation with Simran Khanuja , co-authored paper won best paper at EMNLP 2024!
- Contributed to the open-source multilingual multicultural multimodal LLM project Pangea
Applied Scientist Intern
Amazon Alexa
May 2023 - Aug 2023- Performed data augmentation with synthetic data using generative methods to support speech technology development, particularly with the help of Large Language Models (LLMs), ASR and TTS Models.
- Implemented a pipeline for synthesis and trained, researched, evaluated and analyzed wakeword models.
Research Intern
Artificial Intelligence Institute at University of South Carolina
Jan 2022 - Jun 2022- Developed a BERT based deep learning approach, for multimodal joint embeddings with Prof Amit Sheth
- Worked on a GPT-based medical chatbot and enhanced its capability through context-driven learning.
Research Intern
Wipro AI Labs
Jun 2021 - Dec 2021- Proposed an approach to generate multimodal vision-text joint embeddings using transformer models with Prof. Amitava Das
- Achieved SOTA on meme emotion analysis task and presented the work at AAAI 2022
- Was an Organizing Committee Coordinator for the workshop, DE-FACTIFY, co-located at AAAI 2022.