Deep Learning Performance Engineer
170k - 237k USD
On-site
Full Time
#Engineering
#AI
#Machine Learning
#CUDA
#Systems
#Networking
#Deep Learning
#PyTorch
#Learning
#Ray
#TensorRT
We are looking for a Senior Deep Learning Performance Engineer to join our team on a full-time, on-site basis in the United States. At Anyscale we are commercializing Ray, the open-source framework that powers scalable machine learning workloads for organizations such as OpenAI, Uber, Spotify, Instacart, and Cruise. Backed by Andreessen Horowitz, NEA, and Addition with more than $250 million raised, we give developers and data scientists the ability to move AI applications from a laptop to a production cluster without deep distributed-systems expertise. In this role you will design and deliver performance optimizations that keep our AI infrastructure at the leading edge of speed and cost efficiency.
Responsibilities
- Collaborate closely with product teams to integrate the newest performance improvements into the Anyscale platform, Anyscale Endpoints, and related open-source projects.
- Partner with research groups to advance LLM inference engines such as vLLM and TensorRT-LLM.
- Track emerging techniques across open-source communities and research publications, then implement and extend those best practices within our stack.
Requirements
- Hands-on experience developing or optimizing software that runs on GPUs using CUDA.
- Solid grasp of operating-system or networking fundamentals and a track record of applying those concepts to performance work.
- Practical familiarity with deep-learning concepts and frameworks such as PyTorch.
- Strong English communication skills.
What we offer
The target salary range for this position is $170,112 to $237,000. In addition, the role includes equity participation and a comprehensive benefits package:
- Stock options
- 401k retirement plan
- Healthcare plans with 99 % of premiums covered by Anyscale
- Wellness stipend
- Education stipend
- Paid parental leave
- Flexible time off
- Commute reimbursement
- 100 % coverage of in-office meals
