Building a unified AI-powered conversational chat experience that provides individualized customer assistance, leveraging generative LLM and other AI technologies developed in partnership with Google DeepMind.

Led the next generation Google CRM Infra and Customer Understanding Data Lake.

TLM leading next generation Google CRM

Google CRM- Reporting TL- Infra TL- Founding TL of next generation Google CRM

Google CRM

Led the software architecture team and responsible for enhancing application's architecture to improve stability, security and performance, including- Object oriented, aspect oriented and service oriented architecture- Dependency injection and interception- Logging framework- Scheduler framework using Quartz- Alert framework- Automated unit testing framework- Memcached- Message queue service using MSMQ and ActiveMQ- Performance profiling and optimization using MiniProfiler, AppDynamics and New Relic- Production issue troubleshooting and problem solving- Codebase management and refactoring. Best practices.- Third party software component license provisioningCreated a central development console (AirWatch Dev Console) that was adopted by all .NET developers company-wide. It greatly improved daily productivity by reducing developer's time spent on development environment setup, daily TFS workspace management, code compiling and testing, database deployment, windows service deployment and management.Worked with Java architects to migrate .Net projects to Java.

Responsible for the accuracy of the company's flagship home energy simulation software - EnergyInsights. Successfully brought the software to pass BESTEST and BESTEST-EX from National Renewable Energy Laboratory with high accuracy. Obtained the highest level RESNET HERS rating software accreditation (https://www.power-grid.com/energy-efficiency/apogee-interactive/#gref).Lead developer to extended the web based application to cross platform, supporting -- Mainstream web browsers -- iOS standalone -- Windows standalone -- Embedded Linux -- IVR/ITR -- LiveScribe smart penConducted an EM&V (Evaluation, Measurement and Verification) study using statistics software R with 200K samples and 760K+ data points to reveal the statistical influence and behavior impact of home energy software.Well recognized researcher in home energy industry. Served in the Advisory Committee on Energy Savings of Existing Homes of US DOE and NREL to establish the national standard BESTEST-EX for energy simulation software tools (http://www.nrel.gov/docs/fy11osti/52414.pdf) and in the RESNET Building Performance Labeling Task Force to establish a national existing house labeling standard.

Conducted advanced numerical analysis of an electromagnetic launcher with liquid metal injection and lubrication. Examined potential designs and assisted experimental researchers to determine the best operating conditions.