Hackathons Participated: 15

Qualified to last rounds: 5

Won: 2

Scilab Tool-B0x Hackathon - Indian Institute Of Technology Bombay

Team size: 2 | Type : OPEN SOURCE 

Problem solved:

In computer programming, pandas is a software library written for the Python programming language for data manipulation and analysis. In particular, it offers data structures and operations for manipulating numerical tables and time series. I and my comrade have enabled calls to about 157 functions of Pandas from Scilab. Some of the functions include operations such as applying a function along an axis of the DataFrame, column-wise combining with another DataFrame, returning cross-section from the Series/DataFrame, etc.

*won a cash prize of 15000 INR 

DRUG Discovery Hackathon 2020 - #Innovate4NewDrugs

Team size: 4 | Type : OPEN SOURCE 

Problem solved:     Develop a reinforcement learning-based algorithm to identify lead molecules by emulating ligand-protein interactions

 We developed a reinforcement learning-based algorithm for identifying lead molecules by emulating ligand-protein interactions using Generative Adversarial Neural Networks (GANs).

My contribution to the project was crucial in developing the GAN model. I worked on training the generator network of the GAN to produce molecular structures that closely resembled known ligand and protein structures. I also collaborated with other team members to use the discriminator network of the GAN to distinguish between the generated structures and real structures and ensure that the generator produced high-quality molecular structures.

We used the generated molecular structures to train a reinforcement learning algorithm that determined the best lead molecules for binding to a specific protein. This was done by emulating ligand-protein interactions, which involved analyzing the various characteristics of the protein and the ligand, such as their shape, charge, and polarity.

The reinforcement learning algorithm was trained to find lead molecules that were most likely to bind to the protein of interest, based on the interactions between the generated molecular structures and the protein. The algorithm iteratively generated new lead molecules and evaluated them using various metrics, such as binding affinity and stability.

Overall, our contributions to the project were instrumental in developing a model that had the potential to accelerate drug discovery and development, by reducing the time and resources needed to identify promising lead molecules.

CISCO: Design Thinking Hackathon

Team size: 4 | Type : Product Development & Design development

Problem solved:

Agenda of the hackathon was to provide a software product that can be used to for social causes and in public sector. Proposed an application that will help people to find the spot for parking their vehicle.

Infosys Assist Edge - The Big Bot League

Team size: 4 | Type: Product Development

Developed a flight ticket booking automation bot using edge verve's automation tool and deployed the bot on assist edge servers.