Software Development Manager - ML Compiler, AWS Neuron, Annapurna Labs

Software Development Manager - ML Performance Tooling and Benchmarking, AWS Neuron, Annapurna Labs

The AWS Machine Learning accelerators, including Inferentia and Trainium, deliver best‑in‑class inference and training performance. They are backed by the AWS Neuron Software Development Kit, which optimizes neural network models for these chips.

The Amazon Annapurna Labs team focuses on innovation in silicon and software. It encompasses silicon engineering, hardware design and verification, software and operations, and places emphasis on small, agile, cross‑disciplinary teams.

• Lead a team of compiler engineers to develop, deploy, and scale a compiler targeting AWS Inferentia and Trainium.
• Act as a technical partner with AWS ML services teams, contributing to pre‑silicon design and bringing new products, optimizations, and features to market.
• Leverage deep knowledge of resource management, scheduling, code generation, optimization, and instruction‑architecture development for CPUs, NPUs, GPUs, and other compute units.

• 3+ years of engineering team management experience.
• 6+ years of hands‑on engineering experience.
• 4+ years designing or architecting new and existing systems (design patterns, reliability, and scalability).
• Experience partnering with product or program management teams.
• Strong software design fundamentals and deep understanding of compiler internals (resource management, instruction scheduling, code generation, compute‑graph optimization).

• M.S. or Ph.D. in Computer Science or a related technical field.
• Experience with tool chains (LLVM, GCC) and code‑generation techniques for new hardware.
• Knowledge of compiler internals from front end to run‑time environment with emphasis on AI acceleration.

Location:

Toronto, Canada. Candidates must be currently located or willing to relocate.

Salary range:  –  CAD annually.

Explore the product:

Amazon is an equal‑opportunity employer and does not discriminate on the basis of protected veteran status, disability, or other legally protected status.