• Led development of the MLflow.js from inception to production, establishing code standards and defining roadmap.

• Architected the library using object-oriented principles and design patterns. Implemented modular components and abstractions in TypeScript, creating intuitive APIs while ensuring maintainability and extensibility.

• Engineered Docker with Docker Compose to containerize the backend in Node.js, standardizing dependency management and build processes while streamlining developer onboarding.

• Orchestrated testing and CI/CD pipeline, achieving 87% coverage with Jest integration tests and establishing automated workflows for tests, builds, and releases, ensuring reliable and efficient delivery.

• Built end-to-end machine learning pipelines with TensorFlow.js, demonstrating integration across experimentation and deployment, validating functionality and accelerating adoption.

• Delivered a developer portal using React and Next.js, enabling self-service learning and reducing support overhead.

• Developed end to end machine learning models to optimize user acquisition, retention and personalization.

• Built the analytics platform from the ground up. Presented to C-level executives on the health of the business and provided actionable insights. Initiated and implemented a broad range of projects for business functions including product, marketing, partnership, strategy and finance, etc.

• Created models on customer segmentation and churn prediction, conducted cohort and repeat purchase analysis to drive growth of the business.

• Designed and analyzed A/B testings for the e-commerce site to improve user experience and optimize conversions. Conducted funnel and user behavior analysis to derive hypothesis.

• Performed data mining and text analytics to discover consumer insights on smart devices industry enabling management to make informed decisions. Conducted analysis on audience behavior of video-on-demand and streaming content on the platform.

• Created an image sequence classification model to recognize motions of smart device users. One use case was to enable our smart TVs automatically pause if audience leave.

• Implemented and utilized a wide range of machine learning algorithms and big data processing techniques to build data products.

• Lectured for machine learning and mathematics courses of the M.S. in Data Science program at University of New Haven. Directed Python labs and machine learning practicum.