September 2021 - Recent

Senior Software Engineer

Berry AI Inc.

DevOps

Supporting MLOps Construction and Open-Source Contribution

• Spearheaded the development of MLOps infrastructure, actively contributing to an open-source repository.

• Orchestrated the deployment of over ten on-premise servers and AWS services, specializing in the design and implementation of hybrid cloud CI/CD solutions.

• Designed and implemented a robust diagnosis system to monitor on-premise server status, ensuring optimal performance.

• Successfully led the migration of legacy systems from on-premise to AWS, employing Terraform to establish a streamlined cloud infrastructure as code.

• Achieved a 15% reduction in AI engineers' manual workload by migrating ML project inference jobs to AWS Batch.

• Realized a 20% reduction in deployment time and a 40% decrease in debugging time through the strategic refactoring and tuning of legacy systems.

• Utilized Ansible to effectively manage Edge servers, overseeing a diverse range of services, including system, ML, and software components.

• Demonstrated proficiency in Edge deployment and management services, specifically focused on Ubuntu-Frame. Developed solutions that streamlined Edge computing processes, enhancing stability, including monitoring and self-healing.

Software Development

Surveillance and Real-Time Systems Development

• Spearheaded the development of surveillance and real-time systems using RTSP/FLV, managing multi-on-premise services with 10,000 videos.

• Collaborated with AI engineers to build near-realtime AI products, leveraging a cohesive development approach.

• Significantly improved legacy code test coverage from 0% to 80%, actively sharing effective testing methodologies with team members.

December 2017 - September 2021

Senior Backend Engineer

Ponddy Education Inc.

Software Development

REST API Development and Microservices Architecture

• Spearheaded the development of REST APIs for the Ponddy Reader and various microservices utilizing the Django REST framework.

• Successfully implemented a secure payment system integrating Stripe and WeChat, incorporating async workers and AWS S3 sign signatures. This optimization reduced backend network usage by approximately 99% in file uploads from the frontend, incorporating WebSocket functionality.

• Orchestrated the integration of Canvas LMS with the LTI protocol, enhancing interoperability and functionality.

• Contributed to increased efficiency by building PyPi packages, significantly reducing the time required for new microservice development.

• Developed an Auth application supporting single login, single logout, social login, microservice authentication validation, and JWT integrations, ensuring a seamless and secure authentication experience.

• Executed the creation of approximately 1000 API autotests and 50 end-to-end tests, fortifying the robustness and reliability of the software ecosystem.