I have been intrigued by technology since my early childhood. I believe technology can be used to better our lives. I have been involved in several team projects as well as developed personal projects over the years.
I enjoy participating in hackathons as they challenge me as a developer, teach me collaborative skills and help me learn from fellow developers.
Notable Hackathons:
- 1st place, Digital Health Hackathon organised by SBST VIT Vellore in association with UNSW, Australia.
- 2nd place, VinHack organised by VinnovateIT
Notable Projects:
- Facial Recognition Attendance System (09/2019)
- Visualising and Solving Blocks Problem in AI using Hill Climbing Algorithm (09/2019)
- Dataset Generation Tool (05/2019)
Generate and annotate image dataset for text detection and
object detection using Deep Learning models.
- Sahayak - 1st Prize Digital, Health Hackathon
(12/2018)
TB Reporting tool for informal practitioners in rural areas. Also
acts as central database for TB cases all over India.
- Krishi Sahayak - 2nd Prize, VinHack (04/2019)
Real Time bidding platform for farmers to curb the widespread
underpayment and assistant for farmers which would give
subjective advice based on current soil conditions.
- Web Scraping using Python (08/2019)
Web Scraping website to download multiple images which would
require repetitive actions otherwise.
- Binary Clock (10/2018)
Simple 24 Hour Clock using Binary representation of numbers.
Projects
DoPatti
Decentralised turn based card game using smart contracts on Ethereum.React, Socket.IO, MaticSkills
Python
Java
JavaScript
Flask
Socket.IO
p5.js
Experience
- CereLabs - Intern
May 2019 - June 2019
Interned at CereLabs, Mumbai. CereLabs is an AI, Machine Learning and Deep Learning research and development company. I developed a tool for synthetic data generation which would label the image as well. This helped reduce manual work of labelling. Trained Deep Learning models for object detection and scene text detection (EAST). Wrote a custom algorithm to improve results of object detection model. Augmented document data to improve object detection results. Achieved invoice detection using Natural Language Processing. Garnered appreciation from superiors all the way till CEO for completing tasks efficiently in given time frames.