Telematica (Martian Mobility Inc)
Building next generation of connected mobility
Leading the product and development team building core EV APIs, tools and products to make EV data accessible to businesses.
Backed by YCombinator
A self-motivated and energetic learner with a passion for space tech, sports and startups. I love to solve problems through code, contributing to open source, and indulging in social service.
Building next generation of connected mobility
Leading the product and development team building core EV APIs, tools and products to make EV data accessible to businesses.
Backed by YCombinator
Worked with Learning and Machine Perception (LAMP) team for completing my undergraduate thesis. Presented a thesis on the problem of Continual Learning in the Graph Neural Networks.
𝗘𝗱𝗠𝗮𝗱 is a cohort-based learning platform with premium mentorship to help students ace tech-interviews
EdMad has helped over 1000 students from 15+ colleges across India to transition from a Student to Professional
Our students have job offers from Google, Microsoft, Amazon, Uber and many major tech companies, unicorns and high growth startups.
Official teaching assistant for the following courses. Helped in creating lab material, direct classroom teaching and as a lab assistant.
Continual learning approaches generally address the problem of catastrophic forgetting of data in continuous regular space (mostly for image classification tasks, with images of fixed dimensions). This project explores and extends the idea of Continual Learning to graphs.
Developed a meta-learning(MAML) based approach for classification of chest X-Rays in the NIH dataset. The problem statement is to train a quick learner to classify X-Rays of the chest to one of the 14 common Thorax diseases.
Made a model trained on CLEVR dataset for the course project. The aim was to answer questions asked in natural language referring to various shapes provided in CLEVR dataset.
Used CNNs for image feature extraction and RNNs for question representation, and combined the features through a composite architecture to get the answers.
Member of the Winner team of Alexa India Hackathon 2019. Trained Alexa to act as a music practice partner and provide realtime feedback.
Created a package to automatically generate tests in case of Sanity failure in Global Address Model.
This reduced the developer time wasted in the inefficient method of constructing test instances manually by querying the GAM from >5hours to ~15 mins.
Created a dynamic UAV swarm environment on Robot Operating System and an ad-hoc network using only open source and low cost hardware and software. The dynamic swarm does not rely on a centralized node with ability to add and remove nodes from the network at any time. The paper can be found here.
A Modular, Concurrent and Customizable Protocol Analyzer. Aim was to provide a fast concurrent alternative to Wireshark, which is synchronous in nature.
An active contributor to this open source college project.
Received full scholarship for attending NAAMII Winter School in AI at Pokhara in December 2019. The coursework and lab touched upon various fields in AI and their applications briefly in 11 day span.
Travelled 8000Km in 15 days to meet various local role models and learn about their entrepreneurship journey. The aim was to understand the roots of problems currently faced by India and use technology affordably and efficiently to propose and develop scalable solutions for the same.
Captain of the TT team for :
Mentored a team of 6 managers with a 100% retention rate to invite dignitaries/organizations for talks &
exhibitions to the annual technical fest of BITS-Goa, Quark 2019.
1. Educational
|
2. Competitive Exams
|
3. Sports and Co-Curriculars :
|