Anaconda, Inc.

Strategic Partner Sales Manager

Perle

Innovations in AI data management, visualization, and orchestration

Advisor

Raise Your Bar

AI product & GTM consulting

Check out my work & reviews at https://sheikhshuvo.com/

Founder

Amazon

Senior Business Development Manager 

Allen Institute for Artificial Intelligence (AI2)

I explored concepts in how machine learning models are evaluated, interpreted, and debugged

Entrepreneur In Residence

Voxel51

ML Visualization & Dataset Curation Tools for Camera Sensor Data

Product Strategy and Growth Lead

Mighty AI

(acquired by Uber ATG) 

I joined as employee # 14 and worked with the world's top self-driving car teams to build the training data sets that fuel machine learning models.

Some of my proudest accomplishments include: 

- Leading the pivot of technology and service offerings from general machine learning to automotive perception research
- Exceeding my 2017 quota by 400%+
- Growing a $50K pilot into a $1.2M partnerships within 9 months with a major European OEM
- Leading the strategic relationships with 20+ automotive OEMs, tier-1 suppliers, and tech companies
- Pursuing, architecting, and implementing partnerships with the University of Michigan, the Common Visual Database Foundation, and more
- Speaking at and creating content for major industry events including AutoSens, CVPR, and Automotive Megatrends

Business Development & Partnerships Manager

Techstars

I joined the team as employee # 15 and led the growth of UP Global into an entrepreneurship powerhouse and a thriving community throughout the Middle East and Asia Pacific. 

Some of my proudest moments include: 

- Facilitating and presenting at events across 32 countries, with audiences ranging from 15 to 1,250+ people
- Creating, implementing, and managing a project to pioneer startup events and programming in Iraq and Kurdistan that ultimately reached 2K+ Iraqi youth
- Launching the first remote office in Asia (based in Delhi) - building a team from ground up and exploding reach by 440%
- Developing expansion strategies for emerging markets within Middle East, China, India, and Australia. 
- Consulting with the US State Department in Egypt around developer evangelism and diversity initiatives

Senior Regional Manager, MENA + APAC (UP Global)

Merge Healthcare (now IBM Watson Health)

- Triaged support queries; decreased response times by at least 10% each quarter
- Advocated internally for new features and functionality based on customer feedback
- Created written and video content for customer knowledge base

Client Services Specialist

I build AI teams

I’m a generalist with a wide range of experiences in sales, GTM strategy, and business development across early stage startups to Amazon. 

I focus on applied AI, with a particular interest in computer vision applications. 

Outside of the above, I spend most of my time outdoors with my family (including 2 young girls + dog) and working on my pickleball game : )

Sheikh Shuvo

Member of Technical Staff

LinkedIn

1. Kernel optimization

I lead Liger-Kernel: https://github.com/linkedin/Liger-Kernel, which is the fastest growing open source project from LinkedIn ever. 

I also give a talk in CUDA MODE: https://youtu.be/gWble4FreV4?si=Ry0SPKs41KkTLba0

2. Distributed training optimization

Our optimized zero++ paper got accepted in ICML es-fomo 2024: https://arxiv.org/pdf/2407.01614

3. Inference Optimization

I am the committer in SGLang and lead the optimization around SGLang at Linkedin: https://github.com/sgl-project/sglang/commits?author=ByronHsu

Senior Software Engineer

1. ML training platform (2022/6 ~ 2022/12): 

Scale training cluster to thousands of GPUs on top of Kubernetes.

2. ML orchestration platform (2023/1 ~ 2023/5): 

Built an orchestration layer from scratch on top of Flyte by making significant improvements to Flyte, upstreaming them to open source, and also internally to serve all ML pipelines (data+training).

**Contributed to Flyte, Kuberay, Envd on the side.**

3. LLM training platform (2023/6 ~ ): 

Initiated and designed an LLM training platform on top of thousands of state-of-the-art GPUs, with performant interconnections and fast storage, enabling 100B+ scale pretraining, fine-tuning, and inference.

4. LLM distributed training (2024/1 ~ ):

Speed up internal LLM training by several times with my contribution to DeepSpeed ZeRO++.

**Contributed to DeepSpeed: https://github.com/microsoft/DeepSpeed/commits/master/?author=ByronHsu**

Software Engineer

I work on full stack of ML system:
1. Build GPU clusters on the scale of thousands of GPUs
2. Build efficient distributed and kernel GPU runtimes

I am the committer of @lmsysorg @flyteorg @theASF
You can also find me on Twitter (https://x.com/hsu_byron) or GitHub (https://github.com/ByronHsu)

taplio

The Battle

Sheikh Shuvo

Pin-Lun (Byron) Hsu

Followers

Posts

Engagements

Likes

Ready to grow your LinkedIn presence?