Ryan ( Zun Ren ) Yao
Professional Summary
Experienced software engineer with over 9 years in software development, including expertise in video streaming, codecs, and image/video processing. Skilled in optimising complex systems, delivering scalable real-time solutions, and contributing to backend development and algorithm research. Recent experience focuses on audio processing, recognition algorithm development, machine learning, system design, and microservices. A proactive problem solver and collaborative team player, eager to learn and adapt to emerging technologies.
Professional Experience
- Designed and implemented advanced machine learning models, driving significant accuracy improvements in audio recognition systems through innovative algorithm research
- Optimised Kafka-based recording systems, achieving an 80% reduction in operational costs and a 75% decrease in resource usage while enhancing real-time data throughput
- Led a high-priority proof-of-concept (PoC) for cloud migration of the audio recognition system, delivering a comprehensive performance evaluation within one month, meeting critical business deadlines
- Streamlined the cue sheet generation process by developing robust, maintainable automation tools, cutting manual effort by 80% and improving accuracy
- Upgraded system monitoring capabilities, enabling early outage detection and reducing manual intervention time by 50% through proactive alert systems
- Conducted in-depth client asset analyses, implementing solutions that cut processing costs by up to 80% without compromising reporting accuracy or quality
- Improved multi-profile video encoding speed from 50% to real-time (100%) by integrating hardware accessories
- Designed and implemented a robust live streaming segment push scheduler, delivering seamless integration with AWS KVS in a PoC deployment
- Streamlined live video delivery by integrating RTMP and WebRTC protocols, ensuring high-quality, low-latency streaming for real-time applications
- Engineered a Yoga Pose detection system utilising Mediapipe, enabling precise motion tracking and analysis for fitness applications
- Ensured reliability and performance of live and video encoding systems by maintaining infrastructure, troubleshooting critical issues, and developing new features to meet evolving requirements
- Developed scalable APIs and seamlessly integrated them into streaming systems, enhancing functionality and user experience for video delivery workflows
- Oversaw the end-to-end management of VoD and live streaming billing systems, ensuring accurate invoicing, scalability, and cost efficiency
- Developed perceptual preprocessing filters, reducing bitrate by 60%-90%
- Fine-tuned video encoding parameters and collaborated with the hardware designer for hardware adjustments
- Evaluated spatial/temporal noise reduction to optimise visual quality and bitrate
- Implementing and modifying of spatial noise reduction algorithm to achieve the video 5%~10% bit rate reduced
- Conducted H.264 rate control and adaptive QP evaluation
- Assisted with encoder IPs FPGA verification and quality assessment
Education
Master of Computer Science - National Taiwan University
Bachelor of Computer Science - National Taiwan University
Technical Skills
Programming Languages:
C/C++, Python, Matlab
Knowledge & Technologies:
Linux, Kubernetes, Image/Video processing, Image quality, AVC, HEVC, FPGA validation, AWS, AWS Elemental service, RESTful API, PostgreSQL, MySQL, Machine learning