Delhi, India
I am a versatile individual always looking to enhance my knowledge and interpersonal skills in the fields of Analytics and AI, Data Engineering, Research and Machine Learning through practical applications. I aim to evolve in any given role and implement the best possible solutions in challenging situations. I am passionate about producing music on Ableton in my free time. Learn and Unlearn to Learn is my motto. I am never the obvious choice. I've always been an underdog. Frankly, I've started to own that piece thoroughly. I'm not always the smartest person in the room. But I am the person who navigates through uncertainty and finds the path forward. I don't shy away from difficult problems. I thrive in them. I believe in creating real impact wherever I go. I am adaptable, given the instantly changing situations, which is my power.
1. Cloud Data Engineering & Azure Platform Enhancement: * Engineered a streamlined Azure architecture, consolidating resources and utilizing segregated Data Factories (Ingestion/Transformation) for improved clarity and management. * Developed parameterized REST API pipelines in Azure Data Factory, reducing data ingestion times by 1 hour per job while integrating NetApp CVO for real-time data scanning. * Implemented robust cross-Data Factory pipeline monitoring, ensuring sequential job integrity and preserving complex ETL workflows. 2. Generative AI Accelerator & SDK Development: * Architected and developed reusable accelerators for GenAI workflows, covering data preprocessing, vector embedding, RAG implementations, and evaluation frameworks. * Built a flexible AI gateway wrapper enabling seamless integration with multiple LLM providers through a single interface. * Packaged multiple codebases into SDKs, simplifying integration and accelerating the development of new AI-powered applications. * Demonstrated accelerator value by rapidly developing a multi-format QnA chatbot on Databricks (under 1 week) capable of generating context-aware summaries. 3. Data Modernization & Delta Lake Implementation: * Led the migration of existing Azure Data Factory pipelines from Parquet to Delta Lake format for intermediate and final data layers, enhancing reliability and enabling ACID transactions. * Managed Azure CI/CD pipelines to seamlessly integrate Delta Lake requirements into existing workflows. * Executed the Delta conversion strategy ensuring no disruption to ongoing production processes 4. Operations Research & Finite Schedule Optimization: * Spearheaded the end-to-end implementation (Data Engineering & Data Science) of a Finite Schedule Optimization solution using Google OR-Tools (CP-SAT Solver). * Focused the optimization model on maximizing production output within a defined horizon (<= 3 months), directly addressing core business objectives.
1. Omnichannel Reporting & Analytics (Major Pharma Client): * Engineered the backend Databricks landing layer for omnichannel data, employing predicate pushdown for efficient channel-specific filtering. * Developed aggregated views (User, Regional, National levels) powering Power BI reports to enable a data-driven customer experience strategy. * Managed the end-to-end reporting lifecycle, including documentation, query optimization, report testing (Power BI), and deployment. 2. Machine Learning Model Optimization: * Optimized Matrix Factorization and GRU models using NumPy vectorization techniques, slashing runtime by 42% and enabling timely weekly model execution and validation. 3. Azure Data Platform Engineering & DevOps: * Enhanced a large-scale Azure data ingestion framework (handling 50TB weekly) through ETL improvements, infrastructure upgrades, DevOps practices, and governance implementation. * Improved system stability and reduced Azure Data Factory pipeline failures via robust component design, active participation in code/design reviews, and effective knowledge transfer. * Automated Databricks user access management, cluster configuration, and user/group info extraction using SCIM V2 REST APIs and custom scripts. * Developed a data replication framework (< 1 month) to efficiently sync production data (logs, code, notebooks) to lower environments, accelerating development cycles. * Designed and refactored Azure Data Factory pipelines, eliminating bottlenecks and achieving a 40% reduction in execution time. * Integrated Databricks with SSMS to facilitate operational reporting and data querying. * Implemented PowerShell scripts for Azure Active Directory data extraction needs. * Maintained the central code repository and provided development/design assistance to other teams.
1. Analyzed patient journeys for HIV specialty drugs, identifying critical process bottlenecks; insights directly contributed to initiatives reducing patient dropout by 28%. Developed automated reporting using Python/SQL to facilitate this analysis 2. Executed customer segmentation of the prescriber universe by applying Python analytics to omnichannel data, aligning segments with market research clusters to inform respiratory brand strategy. 3. Developed and implemented a Matrix Factorization recommender system model to predict Healthcare Professional (HCP) attributes, achieving a high degree of accuracy (15% MAPE). 4. Engineered Python automation scripts for weekly sales data processing and report generation, eliminating 32 hours of manual effort each week.
Patient Analytics & ML: Automated Excel reporting to identify HIV patient drop-off bottlenecks, supporting a 28% reduction. Inferred missing HCP attributes via Matrix Factorization with a 15% MAPE.
Reduced user churn by 8% & drove a 40% increase in user traffic & 38% net revenue through data-driven contest design of Fantasy Games for cricket & football.
Automated Excel VBA reports, saving 80% manual effort; mentored 120+ field associates on Microsoft 365 tools.