Harsh in my mother tongue means "happy person".
I teach computers how to learn.
Working my way to understand world. I was born in a family of Farmer. With ambitions, hard work and my ability to learn quickly is the reason I have made it this far. If you want to know more, ask me about a story.
Harsh Patel
1004, E Block, Synchronicity, Powai
Mumbai, India. 400072
+91 - 9790464549
harshpateliitm@gmail.com
Masters in Automotive Engineering and Bachelors in Engineering Design• June 2019
I am in my final year of college pursuing Dual Degree course (B.Tech + M.Tech) in Engineering Design and Masters in
Automotive Engineering. From early on I liked solving problems, so I developed interests for AI, I am currently working with
in MiRL Lab for Applications of Deep Learning and Computer Vision in Medical Imaging.
I did my research thesis on Medical imaging. I used novel convolutional Neural Nets to identify Congenital heart defects (CHD) with accuracy of 92% accuracy, which is 25%+ improvment in benchmark model of SonoNet.
Alpha Research Consultant • Febuary 2019 - June 2019
Netted profits of over $20 million while developing financial models (alphas) as a part of a 900 highly selective global research team. Derived volume, mean reversion, momentum, and sentiment based alphas expressions with a major focus on fundamental and technical alphas. Developed and deployed systematic financial strategies across a variety of asset classes in US, Europe and Asian market, utilizing a proprietary research platform and risk management process through meticulous literature review and back-testing on WEBSIM.
Deep Learning Intern • December 2017 - June 2018
I have worked on 2 major projects during my 7 months internship under Vinay Sudhakaran (Senior Key Expert, Siemens).
One is Prediction algorithms for context aware video frame and path prediction
where I implemented PredNet model which predicts 9 frames using 1 input frame of video with MSE: 0.000605
and improve the prediction framework using probablistic modelling and object tracking.
Second one is Object and anomaly detection frameworks to detect and prevent anomalies
In total, developed 6 modules to create end to end pipeline for computer vision. You can see the results
of the modules here.
Data Analyst (Remote Project) • Octomber 2018 - November 2018
I did a remote project with TokenAnalyst to work on onchain data analysis. I mined and analysed blockchain data to identify patterns in transactions and track movement of crypto assets for various application like exchange volume analysis, whale transaction alerts, etc.
Data Researcher • Febuary 2018 - September 2018
I am working with 3 other IIT Madras Alumni(2013) on this project as part time. I have worked on analysis and evaluation of various cryptocurrency projects (15+) based their cash flow and performance. Here is one of the report for the reference. I have developed various quantitative methods to analyse blockchain data to make better decisions.
Trading Consultant• February 2018 - April 2018
We Implemented Margin Trading feature on crypto exchange for 1st time in India in a span of 2 months which had a huge impact on customer aquisition, trading volume, liquidity and order book spread.
Want to know what I am up to now?