About Me

I am a seasoned Data Scientist with over seven years of experience in the US healthcare and logistics sectors. My expertise spans Machine Learning, Deep Learning, Natural Language Processing, Computer Vision, and team leadership, through which I have made significant contributions to solving complex industry challenges. I have led efforts to develop predictive and statistical models, recommendation systems, NLP solutions, and real-time data processing pipelines, delivering impactful results at companies like BlackBuck and Innovaccer. With a strong academic foundation from IIT Madras and ongoing graduate studies in Artificial Intelligence at San Jose State University, I am passionate about applying AI and machine learning to address real-world problems.

I am currently pursuing a Master's in Artificial Intelligence at San Jose State University , with an expected graduation in December 2025. Prior to this, I earned both a B.Tech in Mechanical Engineering and an M.Tech in Intelligent Manufacturing from the Indian Institute of Technology Madras , graduating in 2016.

Email  /  LinkedIN  /  Github  /  Resume   

Profile Picture
Work Experience

Data Scientist Intern ( Innovaccer, San Francisco)
February 2025 - Current

  • Built an evaluation framework to assess clinical decision-making of LLM agents using real-world patient data, reducing benchmarking time by 80%.
  • Developed an LLM multi-agent system to automate clinical text mapping, reducing data ingestion time by 50%.

    		•

    Sr. Data Scientist ( Blackbuck , Bengaluru )
    April 2022 - July 2023

  • Introduced and deployed a multi-objective recommendation system for trucker-shipper marketplace application to capture user preferences over diversity, relevance, and new shippers by designing a contextual multi-armed bandits model, using the Vowpal Wabbit framework, AWS SageMaker, and AWS Athena.
  • Developed a real-time GPS data outlier detection system capable of detecting 95% of noise data while compromising only 0.1% real pings, by implementing a Dynamic Kalman filter and replicating Python's FilterPy library in Java
  • Improved the NDCG metric of a recommendation system by 3%, by developing a semantic representation of product entities using a GloVe model and utilizing semantics to build features for the recommendation model.
  • Developed a truck availability model using LSTM and Attention on PyTorch with 100k truck's GPS and user activity data, leveraging short and long-term features to predict availability. Achieved 84% accuracy and 81% recall, which improved booking page actions (calls, clicks, bids) by 4%.

    		•

    Technical Lead, Innova Solutions ( Innova Solutions , Chennai )
    September 2021 - March 2022

  • Developed a single-point access data lake of de-identified US healthcare data, coming from multiple data sources using an AWS Athena data pipeline, AWS Lake Formation, and data mesh architecture.
  • Integrated the data lake with AWS SageMaker Studio, Apache Superset, Microsoft Power BI and Tableau to enable teams of data scientists to build ML models and visualizations.

    		•

    Sr. Data Scientist ( Innovaccer , Noida )
    June 2016 - August 2021

  • Led a team of data scientists to develop Patient Identity Management and Risk Management solutions. Helping Healthcare practitioners and payers in population health management in a value-based care setting.
  • Defined and developed an Index to evaluate regional social vulnerability risk using the Principal Component Analysis, to help physicians prioritize patients having the same clinical condition but different socio-economic status.
  • Trained a bidirectional LSTM model, using patient's historical clinical encounters data to predict the onset of congestive heart failure. Achieved an AUC-ROC of 0.85 with the model.
  • Accelerated release time for CMS-HCC, HCC-HCC, and CDPS models from 1 month to 2 days by creating a framework that transformed logical metadata from SAS files into dynamic configurations for the Python codebase, streamlining the entire deployment process.
  • Designed and implemented a microservice architecture with AWS Lambda and PostgreSQL, replacing Spark-based model scoring with Python-based development, reducing deployment time and dependency on data engineering.
  • Spearheaded the development of a Healthcare Data Insights Dashboard on Microsoft Power BI at Innovaccer, delivering actionable patient data insights for Accountable Care Organization (ACO) leadership.

    		•