Staff AI Application Engineer - Heterogeneous Computing
Teknologji Informacioni (IT)Gjithë ShqipërinëDjeFull Time
Position Overview
As a Staff AI Application Engineer in the Model Enablement team (Application Software), you will build our model enablement platform and decide how each model runs across the Metis or Europa AIPU, the host CPU and the host's accelerators. How this work is split and scheduled determines most of the end-to-end performance a customer sees in the Voyager SDK.
The role combines architecture and hands-on work across the model frontend, the graph compiler, the runtime and the application pipeline. During model bring-up, you cover the frontend and graph side, and a device-side engineer covers the backend.
Our team uses AI coding agents in daily work, and you will use them to build, test and verify your own work. We also welcome experienced field application engineers, technical leads and engineering managers who want to return to hands-on engineering. Experience in planning, delegating and reviewing work transfers well to working with AI agents.
Key responsibilities:
- Model enablement platform and agent harness: Build the platform that engineers and AI agents use to bring up models: reusable tools to partition, compile, run and profile; environments with access to real hardware; closed-loop runs that apply a change, measure it against the reference and repeat; and verification (differential testing, regression detection) integrated into CI.
- Model ingestion: Own how a PyTorch or ONNX model is captured, analysed and split across the AIPU's matrix and RISC-V vector units, the host CPU and the host's accelerators (GPU, NPU, video decoder), with placement tags that users can read and override. Build this with the graph compiler team so it works for every model.
- Heterogeneous execution: Extend that split to large models, cascades and multi-model pipelines, for example with expert offloading and caching across device and host memory. Agree the plan with the graph compiler team.