Haibei Zhu

Software Engineer @ Microsoft

About

My motivations are rooted in both entrepreneurial and socially beneficial aims. I believe that profit and social responsibility are not mutually exclusive.

Country

United States

City

Redmond

Industry

Computer Software

Skill

Kubernetes, Azure Kubernetes Service (AKS), C#, Apache Pulsar, Jenkins, Continuous Integration and Continuous Delivery (CI/CD), Apache Spark, Docker, Spring Boot, React.js, React Native, Facebook API, REST APIs, Apache Camel, Amazon Web Services (AWS), Kotlin, TypeScript, PHP, C++, Python

Experience

Software Engineer

Microsoft

2024-8 - Present · 2 yrs 2 mos

Redmond, Washington, United States

Software Engineer

Toast

2023-7 - 2024-8 · 1 yr 2 mos

Fintech: Backend Payment Work on automatic and manual payment capture

Software Engineer

Meta

2022-7 - 2023-7 · 1 yr 1 mo

Menlo Park, California, United States

Full stack engineer Developed a new portal page for a Data For Good case study, utilizing React, Hack, GraphQL, and Relay. This effort resulted in positively impacting an estimated 274,318,500 people worldwide through 29 significant wins, up from 70 million in 2021. Built the FB Researcher API endpoint for DSA article 40, which allowed qualified researchers to access Facebook, IG public post, group, and account data using the Researcher Platform (Jupyterhub). This product is a multi-stack solution involving Meta's search infrastructure, data stores, PHP web layer, AWS, and client-side SDKs in Python and R.

Machine Learning Engineer

Tencent

2021-5 - 2021-8 · 4 mos

WeChat Ads Recommendation ● Implemented ALBERT, SimCSE models for more accurate user advertising, which lifted the month Gross Merchandise Volume up to 0.97%. Optimized feature engineering pipeline with data acquisition (HDFS), and data pre-processing (NumPy, Pandas) for advertisement recommendations ● Built and deployed scalable commodity fingerprint APIs with Spring Boot and Docker to better optimize client advertising.

Machine Learning Engineer

Vanke

2020-11 - 2021-4 · 6 mos

Developed Spark applications for named-entity recognition using multiple programming languages such as Python and Scala to automate the human review process, reducing mistakes and increasing efficiency The LLM paper "Negative Siamese Network for Classifying Semantically Similar Sentences" was chosen for publication at IALP 2021.