COVID-19 Response Group at Centre for Networked Intelligence, IISc.
March, 2020 -- March, 2022
City Scale Agent-Based Simulator (With TIFR, Mumbai)
  • Developed an agent based simulator from scratch using JavaScript that can simulate around 100K agents and runs natively in the browser (One of three members who coded up the simulator engine).
  • Provided support in translating the simulator from JavaScript to C++ for higher performance and scalability. Added new feature like age-based interaction rates in the C++ simulator. The C++ simulator can simulate around 2M agents.
  • Co-wrote Python wrappers and calibration scripts for the simulator. Ran the calibrations in a supercomputing facility at IISc. This was used for running epidemic simulation to predict disease progression at city scale.
  • Softwares: C++, Python, JavaScript, Supercomputing

    Mobility Data Analysis
  • Analysis of mobility patterns to understand travel behavior at large scale using various mobility data sources like Google, Facebook, etc.
  • Designed a local database to store these large datasets (Millions of rows and 10s of GBs in size) and provided access to the rest of the group for ease of information retrieval and computations.
  • Contributed insights from the mobility data. These served as a proxy of interaction rates among communities and/or between different administrative regions.
  • Softwares: Python, PostgreSQL with PostGIS extension

    Workplace Readiness Indicator
  • Designed the score computing algorithm that provides not only a quantitative readiness score but also suggestions on measures to improve the safety of office space so that they could relaunch economic activities in a safe manner.
  • Designed the website and deployed it on an AWS instance.
  • Softwares: JavaScript, HTML, Python, MongoDB, Nginx, Gunicorn

    Swabs2Labs and Optimal Serosurvey Design
  • Developed interactive maps that visualises transport routes from the sample collection centres to the testing labs in Karnataka state. The routes are colour coded based on the number of sample travelling on that route.
  • Deployed both of these tools on the websites, hosted on AWS instances.
  • Softwares: Python, KeplerGL, Nginx, Gunicorn

    COVID-19 Forecast Dashboard
  • A website that displays predicted daily number of cases from different models for all the states in India.
  • Deployed website and setup automatic pipeline for fetching predictions from different models.
  • Softwares: Python, Dash, Plotly
    Website: incovid19/dash.php

    Netradyne Technology India Pvt. Ltd.
    November, 2019 -- March, 2020
    Data Science Intern (Part Time)
  • GeoSpatio-Temporal analysis for finding hidden patterns in driving behaviour.
  • Database management for archiving, handling, and exchange of large data sets.
  • Softwares: Python, PostgreSQL, KeplerGL

    Made with ❤  in IISc, Bangalore.