# Chengzhu Duan

> Software Engineer at Coinbase

Location: San Francisco Bay Area, United States
Profile: https://flows.cv/chengzhu

I am a software engineer who loves building tools that give people easy access to resources and enable them to use their talent to impact the world positively. To do that, I am constantly learning in the fields of computer science, math, economics, and moral philosophy. I am forever an apprentice who loves to collaborate, innovate, and reason.

## Work Experience
### Software Engineer @ Coinbase
Jan 2021 – Present
• Integrated third-party exchange APIs to increase liquidity options for institutional clients using Coinbase Prime, generated $87B+ in total trading volume
• Built scalable and fault-tolerant services to reliably place, track, and recover 10K+ orders per second with data sharding and failover redundancy
• Designed and scaled the market data service to deliver high fidelity price book for 500+ products within 30ms of latency
• Scoped and implemented reusable and extensible framework for novel work to increase developer productivity and code maintainability

### Software Engineer Intern @ Coinbase
Jan 2020 – Jan 2020 | San Francisco, California, United States
• Streamlined an in-app invoice workflow for Coinbase Prime that will save finance and operation team 40 hours of phone call per week, claim back $350K unpaid invoice, and provide 800+ Custody Organizations with timely invoices
• Built scalable backend automatic billing service, invoice service and email notification service with Ruby on Rails
• Ensured fast UI load time with proper frontend caching through Rest Hooks and SWR
• Automated Custody invoice PDF generation and storage in AWS S3 with the use of Rails ERB template and Puppeteer

### Software Engineer Intern @ Omniscience Corporation
Jan 2019 – Jan 2019 | Palo Alto, California
• Expedited image resizing and uploading to AWS S3 by using Thumbnailator to reduce 30% thumbnail load time
• Constructed a scalable RESTful Flask app with Swagger Codegen and SQLAlchemy for reliable underwriting automation
• Implemented document uploading, sharing, editing, and labeling across multiple online platforms with Angular, Spring Boot, and AWS S3 for thousands of underwriters from large insurance companies (e.g. RGA Insurance, AIA Group)

### Undergraduate Student Researcher (Early Research Scholars Program) @ University of California, San Diego - Jacobs School of Engineering
Jan 2018 – Jan 2019 | Greater San Diego Area
• Developed a Bayesian Personalized Binary Classifier that predicts whether or not users will play a game before purchasing from Steam with 93% accuracy
• Utilized Scipy and Scikit-Learn to apply popular visualization and data-processing practices (e.g. t-SNE) on huge real-world data and build custom machine learning pipeline (e.g. BPR with a sigmoid activation function) 
• Optimized the model training with stochastic gradient descent through map-reduce sampling for faster and more reliable evaluation metrics


## Education
### Bachelor of Science - BS in Computer Science
UC San Diego

### High School
St. Joseph's Prep


## Contact & Social
- LinkedIn: https://linkedin.com/in/chengzhu-duan

---
Source: https://flows.cv/chengzhu
JSON Resume: https://flows.cv/chengzhu/resume.json
Last updated: 2026-03-29