Engineering Manager, Hardware Bringup

Modular•United States / Canada,

50d•$216,000 - $372,000•Remote

About The Position

ML developers today face significant friction in taking trained models into deployment. They work in a highly fragmented space, with incomplete and patchwork solutions that require significant performance tuning and non-generalizable, model-specific enhancements. At Modular, we are building the next generation AI platform that will radically improve the way developers build and deploy AI models. As part of our mission to build AI's unified compute layer, we are expanding the Modular software stack to a variety of new and exciting hardware platforms. We are looking for an Engineering Manager to lead the Hardware Bringup team at Modular. In this role, you will own the end-to-end delivery of accelerator bring-up across the Modular stack from Mojo kernels and the graph compiler to MAX model serving across multiple concurrent hardware partnerships.This is a hands on management role: You will work with the HW Bringup architect to set the team roadmap and the technical direction, align stakeholders and ensure all projects are delivered on time. You will participate in design reviews and Pre-sales discussions.. You will be accountable for scoping and delivering against engineering Statements of Work (SOWs) with vendor counterparts, and for growing a high-trust team that consistently ships against ambitious, externally-committed milestones.

Requirements

7+ years of industry experience in compiler engineering, high-performance computing, kernel development, or related domains, with 2+ years managing engineers (strong IC-to-manager transitions welcome).
Strong technical foundation in compilers, systems software, or accelerator programming
Sufficient to be hands-on with the team's codebase, lead architecture discussions, and credibly review designs in MLIR/LLVM, kernel codegen, or runtime systems.
Proficiency in C++ and direct experience working in complex, multi-component software systems.
Hands-on experience with at least one heterogeneous programming model (CUDA, SYCL, OpenCL, or similar), as a contributor rather than only a user.
Demonstrated track record of delivering multi-quarter technical roadmaps with external dependencies, including managing scope, risk, and external commitments.
Experience working directly with hardware vendor engineering teams joint bring-up efforts, partner escalations, or vendor-facing technical leadership.
Experience scoping or negotiating engineering deliverables under formal agreements (SOWs, joint engineering plans, or equivalent contractual frameworks).
Strong written communication

Nice To Haves

Experience leading bring-up on non-GPU accelerators (DSPs, NPUs, custom ASICs, or other novel architectures).
Familiarity with MLIR or LLVM compiler infrastructure at a contributor level.
Background in AI compilers, kernel libraries, or GPU DSLs/DSELs such as Triton, CUTLASS, or CuTe.
Familiarity with how AI operators are implemented at a low level — custom kernels, fused operators, or framework-level integration (e.g., PyTorch at the C++ layer).
Experience drafting, negotiating, or executing engineering SOWs with hardware vendors or cloud partners.
Exposure to model serving or inference optimization workflows at production scale.

Responsibilities

Lead, mentor, and grow a 2 to 7 people team of mixed skillset (compiler, kernel, and runtime engineers) focused on bringing up new accelerator architectures across the Modular stack.
Stay hands-on: participate directly in design reviews, contribute to architecture and prototypes, write and review code for non-trivial portions of the stack (Mojo kernels, MLIR passes, runtime integration), and strive to unblock the hardest technical issues on the team.
Support end-to-end delivery of one or more hardware platform enablement efforts, from initial viability and dev-environment bring-up through merged compiler integration, kernel development, MAX serving, and production-quality model deployment.
Contribute to Statement of Work (SOW) scoping, milestone definition, and acceptance criteria with hardware partners, including translating engineering plans into contractual milestones, negotiating timelines, and managing delivery against commercial commitments.
Partner with Kernels, Graph Compiler, MAX, and Cloud teams to align roadmaps, resolve cross-team dependencies, and produce credible FTE plans and forecasts for executive review.
In collaboration with Head of Mojo, hire to a high bar: define the engineering profile, run interview loops, calibrate the team, and close senior candidates.
Foster a write-things-down, demo-driven engineering culture; ensure the team shares findings on new architectures (ISAs, memory hierarchies, vendor toolchains) through write-ups, internal demos, and onsites.