AI and GenAI Project

- Collaborated with 20+ customers from various industries to understand their technical requirements and propose customized solutions, resulting in increased revenue by 10%.- As a Solution Architect, utilizing tools such as Kubernetes, Docker, and Machine Learning/Deep Learning to develop and implement MLOps solutions that addressed user problems. Also worked with SI partners to plan model drift strategies and identify pain points in daily machine-learning production processes.- Actively participated in AI, cloud, open source, and DevOps community day activities to stay up-to-date on new technologies and trends. Hosted online speaking engagements and wrote technical articles to educate users on MLOps and related technologies.- Provided training to 50+ users on how to use the product and MLOps pipeline to successfully complete AI/ML projects, including building the data and ML pipeline, development environment, and automated machine learning work using the python SDK method.Wrote 100+ partner portal articles and showcase articles to help users and SI partners solve k8s, AI/ML, and MLOps problems independently, reducing the time to explain how to do the configuration and operation by 10%.- Collaborated effectively with cross-functional team members to maintain customer systems and troubleshoot operation needs, resulting in timely and efficient issue resolution.

- As a member of the FarEasTone MA Program, I was responsible for developing digital transformation services for AI, data, and IoT projects, and ensuring project goals were met.- In the interval of the MA Program, Our teams and I try to develop a machine-learning model and service to generate fraud reports that reduced fraud detection by 90%. In this project, I try to understand the telecom domain knowledge and analyze the data to fit the project goal. The project has been publicly praised by the 165 anti-fraud hotlines.- Our team and I try to use IoT sensor data to analyze the Room temperature and humidity status. We build the data ETL pipeline, ML pipeline, CI/CD pipeline, and ML Monitoring. Afterward, These methods are used in the energy management project.- Our team and I try to use cameras to help the government to detect the traffic violation. We use the YOLO-base model to build up the ML pipeline and edge computing. These methods are used in the local government projects in Taiwan and help the team to complete the requirements.

FET MA III Program partner.