Leading a team of GCP Specialist Solution Architects that brings the best-in-class experience for Databricks customers on GCP. Built the stream as a founding team member from Day 0 before launch, through previews and GA. Our team works closely with the field, product, and engineering teams at both Databricks and Google to provide a well architected and seamlessly integrated solution, which is easy to build and manage, so that customers can focus on their business use cases to drive value through Data & AI.

Leading a team of Data & AI solutions architect for enterprise customers, spanning across all industry verticals, using Azure and databricks across Eastern US, Eastern Canada & LATAM.

Leader of technical charter of central AI & Smart Manufacturing team, that helped to scale data science teams and initiatives across 8 different manufacturing facilities in 4 different countries. These teams developed automated mission critical smart manufacturing capabilities and analytical solutions that catered to over 8000 process and equipment engineers.Architected and developed a novel solution that helps to abstract machine learning model building from analytical pipeline. This helps data scientists to be substantially productive and focus on building models in various environments from VMs/Dockers, Hadoop Cluster & GPU.Frequently sought after to guide team members and build their talents to contribute towards solution that is modular, scalable, stable and easy to maintain.

- One of the pioneer of data science at Micron worldwide.- Architected and implemented various hybrid analytical solutions using big data (Spark, Hadoop) and small data (Python, R) tools.- Tested and benchmarked various analytical tools and libraries for custom needs to deliver optimal solutions.- Full stack analytics development: ○ Streaming data from equipment sensors, logs, images and persisting in HDFS, Hbase, Hive, etc. ○ Advanced data analytics including image processing (OpenCV) in Hadoop. ○ Implemented predictive analytics using Python & Spark Machine Learning libraries. ○ Frontend representation of data for descriptive and interactive analytics using javascript libraries such as D3, custom HTML5 canvas visualization, iPython (Jupyter) notebooks, etc. - Other interests include Deep Learning and Graph Analytics.

Designed and developed suite of engineering analytics tool that is used by various groups of research and process engineers. Application used by 1000's of engineers worldwide across Micron network. - Real time data collection from manufacturing equipment tool and sensor logs. - Integrated analytics with time series analysis, variability analysis, prescriptive analytics with auto commonality, statistical process control, etc. - Built interactive descriptive analytic tools with various javascript libraries in canvas and D3 with backend analytics done in R and Python. - Implemented easy to setup and configurable controls and alarms from analytics.

Statistical process control of semiconductor manufacturing by analyzing data, design of experiments, etc. Developed utilities to automate day to day analytics.

Implemented custom analytical tools for machine generated data (ASML photolithography tools) which help both equipment and process teams.

- Developed computational algorithm predicting Cu Surface Morphology Evolution during Electropolishing and its simulation by image analysis in Matlab.- Simulation tool used for research by DOE at Jefferson National Lab for particle accelerator design.Conferences and Publications:- Simulation of Cu Surface Morphology Evolution during Electropolishing, J. Thomas and S.R. Brankovic, 217th Electrochemical Society Meeting, Vancouver, Canada, April 2010.- Computational Algorithm Predicting Surface Morphology Evolution During Electropolishing, J. Thomas, C. Reece, and S.R. Brankovic, 6th SRF Materials Workshop, Tallahassee, FL, Feb. 2010.- Modeling of Copper Surface Morphology Evolution during Electropolishing, J. Thomas and Stanko R. Brankovic at 38th Semi-annual TcSUH Student Symposium, Houston, TX Dec. 2009.- Modeling of Copper Surface Morphology Evolution during Electropolishing, J. Thomas and Stanko R. Brankovic at First International Symposium on Nanotechnology, Energy and Space, Houston, TX Oct. 2009.- Modeling of Cu Surface Morphology Evolution during Electropolishing, J. Thomas and Stanko R. Brankovic at Graduate Research Conference, Houston, TX May 2009.

- Performed statistical analysis on research data and surveys conducted at UH using statistical software (SPSS, etc.).- Chi-square tests, reliability tests, factor analysis, regression analysis, etc.

- Software application development, maintenance and testing at CMMI level 5.- Data analysis using DB2, SQL.