Lead the migration of on-premise processes to Amazon Web Services (AWS) cloud and spearheaded the development of a new business analytics platform. Made critical decisions regarding architecture to ensure scalability and support for geospatial data usage. Architected the data modeling, ETL processes, and ingestion pipelines to seamlessly feed data into the platform. Additionally, developed web scraping processes to gather supplementary data for enhanced analytics capabilities.Key Responsibilities:• Migration to AWS Cloud: - Lead the migration of on-premise processes to Amazon Web Services (AWS) cloud. - Implemented a microservice architecture for a business analytics platform with ECS, Cognito, RDS and Fargate, S3 and EMR for data engineering tasks, and other services like Lambda and Dynamo for ad-hoc use cases.• Business Analytics Platform Development: - Spearheaded the development of a new business analytics platform. - Specially involved in infrastructure, backend and authentication and authorization with tools like FastAPI, Cognito and PostgreSQL. This allowed us to successfully launch the platform and acquire new clients.• Data Architecture and Modeling: - Architected the data modeling, ETL processes, and ingestion pipelines to seamlessly feed data into the platform. - Implemented Airflow for data pipelines orchestration and observability; S3 storage with a bronze, silver, and gold architecture for the data lake; EMR with Apache Sedona+Spark for distributed and scalable compute; and PostgreSQL data model for quick access to insights in the platform. This enabled clients to access up to date information with 3x the speed.• Web Scraping: - Developed web scraping processes to collect supplementary data for enhanced analytics capabilities. - Implemented automation techniques to streamline data gathering processes and improve efficiency.

Conducted research and implemented data center rack temperature prediction models using Python and TensorFlow at National Yang Ming Chiao Tung University.Key Responsibilities:• Research: - Conducted in-depth research on data center rack temperature prediction models. - Explored existing literature and methodologies to inform the development process. - Collaborated with a team of researchers to identify key challenges and potential solutions.• Model Implementation: - Utilized Python and TensorFlow framework to develop predictive models. - Implemented machine learning algorithms and techniques to analyze historical data and forecast rack temperatures. - Optimized model performance through iterative experimentation and fine-tuning parameters.• Data Analysis: - Conducted thorough data analysis to identify patterns and correlations relevant to temperature prediction. - Validated model accuracy and reliability through rigorous testing and validation procedures.• Documentation and Reporting: - Documented research methodologies, model development processes, and experimental results. - Prepared reports and presentations to communicate findings to academic stakeholders.Key Achievements:• Successfully researched and implemented data center rack temperature prediction models using Python and TensorFlow.• Developed robust machine learning models that accurately forecasted rack temperatures, contributing to improved data center efficiency and reliability.

Developed backend functionalities and report generation capabilities for a clustering algorithm product tailored for financial institutions, using Python, Flask, and popular data manipulation libraries like Pandas and Scikit-Learn. Provided support during product launch, initial client installations, and incident resolution.Key Responsibilities and Achievements:• Designed and implemented backend features specific to the clustering algorithm product for financial institutions, leveraging Python, Flask, and data manipulation libraries like Pandas and Scikit-Learn.• Collaborated with frontend developers to ensure seamless integration between frontend and backend components, tailored to the needs of financial clients.• Developed a specialized report generator module to present clustering algorithm results and insights in a format relevant to financial clients.• Provided specialized support during the launch phase of the product, assisting with initial client installations and configurations.• Demonstrated proactive incident resolution capabilities, leveraging expertise in Python, Flask, Pandas, and Scikit-Learn to minimize downtime and optimize product performance to meet the clients' needs.