William Zhang

• Incoming Software Engineering Intern for Fall 2026

• Reduced manual qualification effort by 80% by designing event-driven workflows in Go with back-pressure, idempotent retries, and failure isolation
• Cut firmware qualification time from 1-2 weeks to 1 day by designing a no-touch firmware validation pipeline for 1,400+ Bering libraries
• Maintained 99.95% data-path availability by automating drive thermal checks, DCO tests, and movement validation across 2,800+ Glacier libraries

• Increased video analysis throughput by 6x (from 180s to 30s) by parallelizing preprocessing and feature extraction with bounded worker pools
• Built a distributed task pipeline to process 10,000+ videos/week using async I/O, multiprocessing, and compact serialization
• Deployed containerized microservices with Docker and optimized CI/CD workflows to reduce build time from 10 minutes to under 3 minutes

• Developed a low-latency speech-to-text engine using a fine-tuned Soniox Transformer, improving transcription accuracy to 98.3%
• Implemented a real-time streaming pipeline in Python and C++ with buffered concurrency, reducing end-to-end latency from 5s to 0.9s
• Containerized the full pipeline on AWS ECS with autoscaling and fault-tolerant health checks, maintaining 99.9% uptime

Honors: Engineering Dean's List, University Honors

Relevant Coursework: Data Structures & Algorithms, Web Systems, Software Engineering, Machine Learning, Computer Vision, Computer Security