Yihan Zhu

Designed and developed SnowOver Flow, a social media app for the snowboarding and skiing community, providing features such as trip planning, buddy matching, carpooling, real-time chat, weather forecast, and resort information. Built and deployed backend microservices supporting real-time data pipelines using Flutter, Dart and Firebase to synchronize resort, chat, and weather data across 3000+ ski resorts worldwide, ensuring low-latency updates. Scaled the platform from 500+ beta testers to 2,000+ active users in 2 months, with 400+ trips posted and 95% positive feedback. Delivered the MVP with a team of 4, utilizing Xcode Cloud CI/CD pipelines to maintain 99.9% uptime, streamline deployment, and reduce app build times from 1 hour to 10 minutes.

Contributed to the Zaion IO team, facilitating the integration and orchestration of Zaion’s industry-leading SaaS solutions for core banking modernization and transformation, utilized by Wells Fargo, HSBC, ING, CIBC, and US Bank. Built a scalable Kubernetes and Docker-based distributed microservices environment with Terraform for infrastructure as code and Jenkins for automation, reducing cloud costs by $400K annually. Designed a CLI tool with Go and Cobra to streamline the daily use of the microservice environment, reducing setup time from 3 hours to 5 minutes and improving productivity for 100+ engineers across teams. Led a team of 3 engineers to develop and maintain streaming ETL pipelines using NiFi, Flink, and Redpanda, ensuring 99.999% data accuracy over 1TB of weekly data, contributing to the company’s acquisition.

Optimized graphics drivers for Snapdragon chips using C and C++, enabling smooth multimedia playback and advanced rendering capabilities for automotive manufacturers, reducing system latency by 10%. Enhanced rendering pipelines using OpenGL and DirectX, optimizing shaders and reducing frame processing times from 20ms to 17ms, resulting in improved responsiveness and visual quality in automotive infotainment systems. Delivered production-ready driver features, utilizing GDB for debugging and Perf for performance profiling, reducing feature deployment timelines by 1 week while ensuring system stability across hardware configurations.