AI Systems Architect

The Opportunity

This is a founding AI Systems Architect role at the forefront of the next wave of embodied AI. Joining a well-funded startup that has already delivered over 33 million chips, you will be the first hire in this function, reporting directly to the CDO. You will possess high visibility and massive influence, shaping the next-gen compute platforms that bridge the gap between mathematical machine learning models and custom silicon.

Key Responsibilities

• Workload Analysis: Analyse modern AI model architectures, including transformers and emerging alternatives, to determine their computational and system requirements.
• Hardware-Software Co-Design: Work closely with hardware architects to translate complex model requirements and computation graphs into efficient silicon implementations and dataflows.
• Model Optimisation: Identify opportunities to modify or optimise algorithms and dataflows (e.g., quantization, pruning, sparsity) to improve performance, power efficiency, and scalability on hardware.
• Architectural Modeling: Develop performance and efficiency models to guide hardware design decisions for both training and inference pipelines.
• Frontier Research: Stay current with emerging AI research to identify when new architectures (such as world models or continuous-time networks) require specialized hardware support.

Candidate Profile

We are looking for an interdisciplinary expert who enjoys solving problems at the boundary of machine learning, software frameworks, and hardware architecture.

• Experience: Minimum 5–7 years of experience specifically mapping AI models onto hardware.
• Industry Background: Proven experience at a top-tier AI lab or hardware company (e.g., NVIDIA, Meta).
• Technical Expertise:
  • Deep knowledge of Transformer architectures and related model families.
  • Familiarity with VLA (Vision Language Action Models), VLM (Vision Language Models), and ViT (Vision Transformers).
  • Understanding of NPU/SoC design and the hardware constraints involved in chip-level implementation.
• Frameworks: Expert-level proficiency with PyTorch, JAX, or TensorFlow.
• Education: A PhD with hands-on, real-world implementation experience is strongly preferred.
• Mindset: A high-influence individual comfortable shaping technical direction as a founding team member.

Logistics & Compensation

• Salary: $300,000 – $325,000.
• Equity: Generous founding-team equity package.
• Location: Santa Clara, CA. This is an onsite/hybrid role; candidates must be able to work in the Bay Area.
• Sponsorship: H1B transfers and visa sponsorship are available.