Software Engineer Neuron Containers, Annapurna Labs, Neuron Containers, Annapurna Labs

Annapurna Labs was a startup acquired by AWS in 2015 and is now fully integrated. AWS Neuron is the complete software stack for the Inferentia and Trainium ML accelerators — the custom silicon that powers large-scale AI workloads on AWS.

The Neuron Containers team is looking for a Software Development Engineer to build platform integrations that enable customers to run distributed training and inference workloads on Neuron at scale. The team owns Neuron integration with Kubernetes, ECS, and Slurm — handling device allocation, fault tolerance, auto-scaling, and orchestration across large clusters. The team also owns delivery of Neuron Deep Learning Containers (DLCs) and Deep Learning AMIs (DLAMIs) — the pre-configured container images and machine images that package the Neuron SDK for customer deployment on Trainium and Inferentia instances.

Key job responsibilities Design and implement container platform integrations — device plugins, DRA drivers, and operators for ML accelerator resource management Build and maintain Neuron DLCs and DLAMIs for customer deployment across EKS, ECS, EC2, and SageMaker Diagnose and resolve performance and scalability issues across large customer clusters Simplify systems — deprecate legacy software and reduce complexity in container delivery pipelines Deliver software across the full development lifecycle including design documentation, implementation, testing, deployment, and operations

A day in the life You'll work with teams across the Neuron org and customers to build and maintain integrations for current and next-generation accelerators. You'll participate in architecture reviews, triage test failures, resolve operational issues, and contribute to upstream Kubernetes projects. You'll debug platform integration problems — how Neuron interacts with container runtimes, orchestrators, and scheduling systems a scale.