Weifeng(Brian) Lai Email & Phone Number

Durham, Nc, Us

Durham, North Carolina, United States

Work as a SDE in Disaster Recovery (DR) Core team, with a focus on snapshot replication (Async, NearSync, Sync).• Designing and developing advanced replication engine for DR workflows such as synchronous replication, virtual disk relinking, and distributed write mechanism to rocket replication performance and resolve data integrity issues.• Boosting the performance of snapshot comparison algorithm by redesigning efficient interactions with distributed metadata databases through self-designed multi-level caching. Leveraging modern C++ techniques like move semantics to squeeze out better performance on cache management. Serving as a foundational workflow for all DR services.• Optimizing an efficient capability check workflow with well-designed extensibility for the replication backend in multi-node and multi-cluster scenarios. Serveing as a cornerstone for all DR services.• Developing data integrity checker and stat tools for enormous snapshot based on high-performance data structures.• Participating in and contributing to scaling-out architecture design.

Boston, Massachusetts, United States

Assisted Professor Dino Konstantopoulos in INFO 6105 Data Science Engineering Methods and Tools.

San Jose, California, United States

Worked as a software engineer intern in disaster recovery team, Nutanix.• Improved observability of DR core distributed component Cerebro for key projects such as Sanpshot2S3 and CloudDR by C++17. Designed and developed state latency stat of finite state machine, and data traffic monitor during snapshot replication. Enabled Dev/QA teams to explicitly track data replication errors, drill down to the time delay of a specific state of the state machine, the type of replication for each operation, and multiple data metrics in a particular replication file.• Preserve byte-related information into Write Ahead Log under multiple replication workflows for failback and failover when Cerebro node crashes and restarts.• Designed and developed project Cerebro Observability in order to revamp Cerebro stat pages that have grown over decade. Redesign these pages giving it a logical structure and rich information that enables Dev/QA to debug issues quickly. Won best team Award in U-Hack 2022, which is Nutanix’s hackathon for all interns.

Shenzhen, China

Worked as a Back-end SDE in Tencent YouTu AI Lab (Top Tier Computer Vision Lab in China), Tencent Cloud.• Developed and optimized monitor service of YouTu Nginx, which is the access layer of Youtu AI service platform, via combining authentication module, CMDB configuration platform and L5 load balancing module, etc. Visualized log information for particular OCR service and provided data support for automated search for predicted failure images.• Provided company information correction support for OCR result. Prepared data, and implemented web service with high performance crawler, Tornado, ES, TRPC. Optimized search time by 60% via switching ES index analyzer, pruning keyword, flush interval, and ASync IO. Increased recall of OCR services of corporate Information by 3%.• Streamlined web service deployment process of AI model by designing and developing combination strategies, such as reusable web server, model resource configuration and initialization, execution handler, GPU binding strategy. Improved security when delivering models via tools such as model encryption, license authorization, json-proto, and py2so. Completed scripts and docker files to wrapping model and service as SDK & image for efficient storage and utilization. Saved 80% of the time that development team spent to deliver services to execution team.• Designed and developed OFD electronic document parser for extracting and rendering coordinates of objects, words, lines, and pictures from invoice, announcement and other e-document via zip, freetype, cairo, opencv and other libraries. Expanded input options and visualization capabilities for the electronic information flow of the company’s OCR business.

Shanghai, China

Worked as a software engineer/community developer in open source project Milvus, which is vector database built for similarity search, and it has 26.5K+ stars in GitHub (04/13/2024).• Implemented S3 storage module for Milvus based on AWS S3 SDK, MinIO, and modern C++ 17.• Designed, developed and modified general storage module based on polymorphism in order to support different storage strategies, such as local disk, S3 and HDFS.

Master Of Science - Ms, Software Engineering Systems

Major in Software Engineering Systems; Grad TA in Data Science

Bachelor Of Engineering - Be, Software Engineering

Member of Xiyou Linux Group, an opensource community with a foucs on Linux Kernel; Member of Xiaoge Li NLP Research Group; Major.

Education record

Summer School