Qing Lan Email & Phone Number

Mountain View, Ca, Us

Mountain View, Ca, Us

Sunnyvale, Ca, Us

As the Engineering Manager for AI/ML at LinkedIn, I played a pivotal role in building and developing a high-performing team to tackle complex challenges in AI technology. My responsibilities included leading the implementation of innovative solutions such as the Deep Learning Framework (PyTorch) and vLLM, SGLang and TensorRT-LLM for efficient inference processing. I also focused on optimizing the AI Platform to support a large number of GPUs for diverse GenAI use cases, ensuring seamless resource management and runtime services.

Seattle, Wa, Us

In my role as a Tech Lead Manager at Amazon Web Services (AWS) in the San Francisco Bay Area, I led a team to develop inference solutions for Large Language Models (LLM). I was instrumental in creating AWS's first Large Model Inference (LMI) container solution, which has been widely adopted in the industry.

Seattle, Wa, Us

Tech Lead of Machine Learning inference optimization, large (language) model support.Leading the team to Build Model partition/inference solution across multi-gpu hosts with different MP strategy. Achieving very low latency/high throughput.

Seattle, Wa, Us

Team lead in Sponsored Product Advertising of Amazon.Billion-parameter model launch: the first large-scale model in the Advertising domain of Amazon. With advanced multi-threaded inference and caching system to achieve very low latency.Real-time signal collection system: a service to capture 150k/s logs per host across all hosts in NA, EU and Asia. It provides less than 200ms latency on log processing and less than 1ms P99 latency to the system. The service now served as the foundation of online learning infra for the next generation Sponsored Product models.

Seattle, Wa, Us

Co-Author of DJL: https://DJL.aiFocus on Deep Learning distributed training and high performance inference. Helping internal teams and external customers building up high performance ML infrastructure.Foundation of AI/ML infrastructure, SageMaker Training platform.AWS MXNet committer.Elastic Inference Accelerator

Santa Clara, Ca, Us

- Duplicate-Bug Recommendation Services: Expected to save $600,000 per year and pushed into production (NVBugs)- Automated pipeline: Data Collecting - Training - Deploying- Top Contributor: Awarded Top Contributor in the team in Midterm and Final Evaluation

New York, Ny, Us

Team Leader of Intelligence Network Switch on Android. Leading Prof: Henning SchulzrinneUsing Location based Networks Information and Application Data Traffic to create Decision on Switch/No-Switch to current Networks.Data cleaning on hardware information. Building stable data management backend using Amazon RDS.Building Decision Model to decide whether the Network should be Switched

Master’S Degree, Computer Engineering

Master Of Business Administration - Mba, Business/Commerce, General

Bachelor’S Degree, Mechatronics And Robotic System

Non Degree, Electrical And Electronics Engineering