• Led a cross-functional team of Data Scientists and developers to design and deploy cost-effective, secure Generative AI solutions for BFSI, Hospitality and Retail clients•Developed AI products, including a Business Intelligence platform for natural language database queries, a resume parserbot on Azure, and an audio analytics solution for insurance QA for Insurance client•Implemented a multilingual text parser for data enrichment, improving the quality and depth of data analysis for Retail and Hospitality clients•Utilized technologies such as multiple LLMs, LangChain, Agentic Workflow, Quantization, and Guardrails to build robust AI solutions aligned with security and compliance standards

Developed comprehensive dashboards to analyze and visualize key metrics, enhancing data-driven decision-making for TD Bank's co-branded credit card partnership with NordstromDevised and implemented innovative methods to bolster customer acquisition and optimize pricing strategies, resulting in significant revenue growth and enhanced profitability for TD Bank's co-branded credit card offerings with Nordstrom

Improved automated chat response system for Indigo Airlines, utilizing OpenAI's GPT-3 model. Tested the model and got approvals for production deployment within a weekUtilized Azure AI services to create a Resume Shortlisting Application for HR

MedIn: Live Inventory Prediction Solution•Built a venture-funded SaaS business and enabled product-led growth by managing a team of 6 people consisting ofdata scientists, software engineers•Designed, developed machine learning model to predict live medicine demand for pharmacies in India•Utilized Zoho Sprints/Jira to create project plans, define project roles & responsibilities, perform project tracking

Early Warning System•Led the entire data science lifecycle, including defining business objectives, collecting and preparing data, building and deploying models, and evaluating performance•Extracted data using SQL queries, command prompt, putty & stored data in databases such as MySQL, Json, Parquet•Cleaned data and performed EDA on data using Numpy, Pandas, Matplotlib, Seaborn, PySpark, Power BI•Analyzed data using Hypothesis testing and Machine Learning Models such as DeepAR, Isolation Forest, Z-Score•Implemented a DeepAR model to predict anomalous calls from healthcare Providers, reducing call costs by 20%

•Call Prediction Analysis– Target: Prediction of probabilities of calls from Providers in a given period of time– Models Explored: Machine Learning models such as XG-Boost, Random Forest, SVM, Logistic Regression– Technologies Utilized: SQL, Python, H2O(AWS), Putty, DBeaver, Apache Hive– Achievements: Created a trained model that can predict call probabilities with significantly improved Precision and recall than previous builds– Deployed the model after end-to-end testing, automated the deployment using Python•Provider Issue Detection– Target: Analyzing and figuring out issues faced by providers and solving them beforehand– Project required heavy cleaning, understanding, exploration, visualization of a large corpus of Data– Technologies Utilized: SQL, Python, Putty, DBeaver, ApacheHive– Achievements: Tagged reasons for various Provider related issues and created value stories out of them•Implemented and extracted reports of personalization on various websites using Adobe Target•Implemented Adobe Analytics on websites, extracted data from warehouse, created meaningful BI reports out of data•Used SessionCam, a tool to record user experience on the web-site, to created funnels, heat maps of user activity•Created code for web scraping, using Python, to extract medicinal data from various websites