• Are you open to working remotely? Yes
  • Please Enter All Locations You Are Open to Work (each line will be each city, country location)
    United States
  • Allow Profile Promotion To Recruiters/Companies Yes
  • Allow Profile Promotion on Social Media Yes
  • Allow Profile Promotion to Alumni Yes
  • Viewed 533

About me



  • 2022 - Present
    nuVizz Inc.

    Data Science Intern

    • Working on developing e-commerce warehouse clusters using multidimensional graph and density based clustering algorithms to find the optimal path to pick orders.

    • Devised a pipeline using AWS and Spacy to extract invoice information from PDF documents.

    • Processed raw documents using AWS Textract to obtain raw invoice information with a confidence level of 97-99%.

    • Developed a named entity recognition model using Spacy to recognize custom tags consisting entities from the invoice preprocessed.

  • 2020 - 2019
    Robert Bosch Centre for DS and AI

    Project Associate

    • Developed a dashboard using Dash and Plotly to display real-time operations of ambulance five districts in Tamil Nadu, India.

    • Developed a clustering algorithm based on DBSCAN to identify the incident locations of an emergency request placed to the EMS system which identified the exact coordinates with an accuracy of 87% over five precision levels.

    • Developed a static ambulance allocation algorithm using optimization of a custom loss function. Improved the area covered by the ambulances with 13% lesser ambulances.

    • Constructed epidemiology models to track the spread and analyze the interventions during the COVID 19 pandemic for State Government of Tamil Nadu.

  • 2020 - 2022

    Lead ML Engineer

    • Led the modelling team on developing a named entity recognition model using Spacy, BERT and LSTM models to recognize deed and gazatte document information. Model gave an accuracy of 92%.

    • Collaborated with data collection team to extract images from Sentinal 1, 2 and 2A.

    • Worked on a semantic segmentation model to access climate risk factors on flood prone areas satellite images datasets with an accuracy of about 89.5% was developed.