• Developed scalable backend contextual recommendation API leveraging EBR and LLaMA-based models for Facebook Marketplace Search team. Utilized user commerce intents and search queries to generate product-to-query, user-to-query, and product-to-product suggestions using model deployment and data pipelines.

• Designed and deployed a real-time asynchronous architecture using Memcache to serve product-to-query recommendations at global scale. Launched as a new module in the Marketplace product details page, driving a 3.4M increase in daily search sessions worldwide.

• Drove cross-team collaboration and data-driven prioritization by conducting extensive SQL/Presto analysis for headroom planning and opportunity sizing. Built dashboards to track impressions, clicks, and CTR of A/B tests, driving highly deterministic business impact with measurable outcomes.

• Led the development of a cutting-edge Large Language Model-based (LLM) Product Recommendation System for an Automobile Parts E-Commerce website. Leveraged HuggingFace Transformer libraries and AWS Sagemaker API endpoint inference service to create highly accurate product recommendations based on the trained AI model's inference. Implemented a REST API layer/endpoint to integrate model output data into the existing webstore API architecture.

• Developed an ETL pipeline to update inventory data and monitor customer cancellations of valid orders. Achieved a remarkable 20% reduction in order cancellations and a 12% increase in GMV revenue within a single quarter.

• Established and supervised project charters/proposals, defined OKRs, and ensured alignment across multiple teams to achieve milestones within an Agile framework. Collaborated with stakeholder leadership to prioritize projects based on their strategic value and business impact.

• Designed and implemented over 10 features in C++, C#, and Python to drive OneDrive/M365 premium subscriptions using an A/B testing framework. These include user experience enhancements in Windows/MacOS applications, XML-based notifications based on product and quota usage.

• Pioneered an innovative project resulting in a 33% increase in premium subscriptions by incentivizing users to back up/sync camera roll photos to OneDrive. Employed object-oriented design patterns, streamlined UX/UI workflows, conducted end-to-end unit testing, and established a robust logging framework.

• Developed an automated PowerBI-SQL pipeline for server-side telemetry data ingestion, facilitating insightful visualizations for business and engineering stakeholders across multiple OneDrive projects. Orchestrated ELT/ETL data pipelines to extract information from a NoSQL database, enabling the generation of actionable reports, and expediting decision-making processes by 40%.

• Served as the subject matter expert (SME) in Robotics Applications, to increase technology adoption, provide technical guidance and recommend applications for Collaborative Industrial Robots, AGV, and Vision systems. Played a pivotal role as an architect during the design phase, serving as a partner in developing solutions, conducting code reviews, and testing to ensure the deployment of robust solutions.

• Led a cross-functional team of 5 software engineers accountable for servicing large automotive accounts. Delivered over 25+ projects within a two-year period, yielding a remarkable 110% increase in revenue across two geographical markets.

• Devised and executed quarterly/annual Product Adoption Roadmap for the company, achieving an exceptional 95% on-time delivery rate for end-to-end software projects.

• Directed recruitment initiatives, conducted comprehensive team member training sessions, facilitated bi-annual goal setting, and executed performance reviews. Efficiently managed long-term project assignments to steer the accomplishment of roadmap targets.

• Amplified visibility for individual contributors (ICs) and provided technical mentorship throughout the software product lifecycle.

Developed a mobile sensor fusion system in C++ using a NVIDIA Tegra X1 processor, a monocular camera with inbuilt accelerometer and gyroscope and a GPS sensor to localize any mobile system (humans, drones or vehicles) in real time on Google Maps in GPS devoid regions (a DARPA ASPN project - All Source Positioning and Navigation)