Research Scientist, Shaip AI, Louisville, US

Nov 2022 - June 2023

•  working on patents and research papers in the clinical domain

•  Filed 2 US patents on clinical NLP and LLMs for clinical domain (under review)

Software Engineer, NLP Cactus Communications, India

May 2022 - September 2022

• Developed & Deployed a rule-based sub module for an editing tool used by researchers to check consistency

Data Scientist I, Verisk Analytics (Remote)

September 2021 - April 2022

• Experimented on boolean question & answering model using SOTA (T5-11b) and other language models to reduce manual efforts; accuracy 87%

• Developed a reusable code for building 100+ text models using keyword matching for underwriting; accuracy 92%

• Analyzed & validated insurance data quality from multiple data sources in different formats using python libraries

Product & Engineering Intern, Apptio India

June 2020 - November 2020

• Developed an interface for AWS deployment tasks for multiple stakeholders; saving 2000+ developer man-hours per year

• Forecasted resource utilization of AWS account for next 2 weeks using machine learning; enabling resource plan

Full Stack Data Scientist Trainee, PixiuAI India

June 2020 - November 2020

• Founding member; launched MVP to find relevant finance news & conversation in the market to ascertain stock value

• Built customized news feed using a multi-class text classifier with a precision of 87%; increasing user retention by 22%

• Designed stocks dashboard by scrapping multiple news sources & applied sentiment analysis; used by 500+ users

Data Analyst, Civilsdaily India

January 2019 - May 2020

• Leveraged text modeling techniques for the UPSC Ed-tech startup with 100k+ online visitors (MAUs) & 50k+ paid users

• Identified novelty of multiple question papers by implementing textual similarity model with 82% accuracy

• Extracted customized entities for a given text using a custom entity recognition model & AWS Comprehend

• Analyzed user behavior using google analytics to understand product performance in 2 tier and 3 tier cities