Palkin Jangra

Working on building a GenAI recommendation system that leverages Machine Learning and Language models technology to provide user-centric recommendations which include understanding of product architecture, vectorization using langchain, implementing and building AutoML tools, API designs, exploring different LLM’s and integrating their API’s. This project helped to raise 1.1 million euros in pre-seed funding. Developed a tool for efficient identification and evaluation of the Corporate Digital Responsibility relevance of companies. This included automation of text data processing, calculation of CDR scores and dashboarding. This tool analyzed over 100,000 documents and reduced manual processing time by 70%.

Worked on identification of duplicate losses submitted by clients based on multiple conditions using PySpark, improving pricing accuracy for actuaries and underwriters by 20%, and developed dashboards in Power BI using cleaned data for the conscious assessment of risk, reducing assessment time by 30%. Developed an end-to-end Machine Learning pipeline to classify loss documents into categories using Natural language processing and tree classification techniques, increasing classification accuracy by 27%. Formed access groups for specific projects in Palantir, ensuring compliance with data governance policies by collaborating with the data governance team to understand data pipelines, branches, and data lineages, improving data security and access management. Created a data wrangling library in Palantir, automating dataset processing and generating statistical reports using Python and ReportLab, reducing processing time by 40%.