About Me

Hello! 🔗

The following is a short introduction thats meant to give a general info about me to the people that stumble on this site.

I’m working as a Deep Learning Engineer in Medical-AI, after recently graduating with my bachelor’s in Computer Science from Amrita University, Amritapuri. I continue to reside in Amritapuri, a serene coastal town where the beaches, lush greenery, and nearby ashram create an inspiring environment.

During my undergraduate studies, I spent time with Neural Networks and PyTorch when working and linear algebra when studying. This led me to work on many interesting research and open-source projects in AI, including a contribution to DeepChem as part of Google Summer of Code 2024.

I read a lot of books, blogs, papers and am always looking for new things to learn — checkout my Bookshelf to see what I’ve been reading lately. I also play table tennis, swim, kayak—and I’m an avid fan of Wes Anderson movies. Apart from English, I know Malayalam and am currently learning Hindi using Duolingo and by talking to friends. Talk to me about Deep Learning(AI), Tech Stuff, Philosophy, Movies, Books, Dogs, or anything that you find interesting. I’m always up for a good conversation.

I’m was active member and mentor of the student-run FOSS club, amFOSS, where I stayed back after classes to work on my projects, particpate for hackathons and organise events. Being a part of this club helped me get exposed to the tech world and allowed me to meet and interact with a lot of amazing people.

If you are interested in the work I do, check out my GitHub profile https://github.com/gauthamk02 or my CV. You can reach out to me at [email protected]. Also follow me on Twitter @gaushn_ to read my random thougths and rants.

My Works 🔗

Member of the EUSML project that works on building AI-assissted tools for the Endoscopic Ultrasound(EUS) procedure. EUS is a complicated procedure that takes a lot of time to master and we collect annotated data of the procedure from experienced doctors and are building an AI model that can detect the stomach stations accurately in real-time during the procedure. My responsibilities in the project include maintaining the dataset generation pipeline, model training/evaluation, researching new & existing methods, and maintaining the codebase.

DeepChem is a Python library for machine learning and deep learning on molecular and quantum datasets. I contributed to the library as part of GSoC 2024 for the project Torch Compile and Pytorch 2.0. You can chckout the Project Report for the technical details of the project. I have also made significant contributions to the library in the past including porting Progrssive Multitask Model from Tensorflow to PyTorch, fixing bugs etc.

Worked on the ai-economist project under Tech4Good Lab that uses multi-agent reinforcement learning for modelling apprenticeship learning. The project constructs diferent scenarios of apprenticeship, mentoring, government policies and simulate them using RL to find and interpret the results.

Researched different methods of Sample Size Determination to predict the sample size required to train a Machine Learning model to achieve a certain accuracy and explored various algorithms and evaluated their effectiveness in different scenarios. As part of the research I also explored McNemar’s and other statistical tests for determining the better-performing model.
Report

Timeline 🔗

2024 🔗

• ➔ Joined TeXNano as a Research Intern

  Working on building AI-assisted tools for the Endoscopic Ultrasound(EUS) procedure as part of EUS-ML project.

• ➔ Got Selected for Google Summer of Code with DeepChem

  Integrated torch.compile into the library over the summer as part of the program.

• ➔ Attended GitHub Field Day - Hyderabad

  A one day conference bringing together technical student community leaders

• ➔ Gave a talk on the Transformer Architecture

  Presented a talk on the working of LLMs and the Transformer architecture in FOSS United Kochi Meetup.

2023 🔗

• ➔ Research Team Member at Tech4Good Lab, UCSC

  Worked on the ai-economist project under Tech4Good Lab.

• ➔ Started contributing to DeepChem

  Ported DeepChem's Progressive Multitask Model from Tensorflow to PyTorch.

• ➔ Won Hackverse 4.0

  First prize winners of the hackathon, Hackverse 4.0, organsied by NITK Surathkal and Devfolio.

• ➔ Joined EUS-ML project

  First major project undertaking.

2022 🔗

• ➔ ACM Internship

  Completed a 2-month internship with and wrote a report on Sample Size Determination.

• ➔ First Hackathon

  Participated in my first offline hackathon.

2021 🔗

• ➔ Joined amFOSS

  amFOSS is one of India's Leading Student-run FOSS clubs.

• ➔ Joined Amrita Vishwa Vidyapeetham

  Started my undergrad in Computer Science with specialization in AI.