Machine Learning Engineer

Brain Chip is looking for a Senior Machine Learning Engineer to build and deliver state-of-the-art models that run on low-power inference hardware, including our Akida Neuromorphic System on Chip.

This is a research and delivery role.

That distinction matters. We're not looking for someone who hands off equations for others to implement. You'll develop novel model architectures, write clean and well-structured Python and PyTorch, and partner directly with hardware and compiler teams so your models run efficiently on silicon. The rigor of your ideas and the quality of what you actually ship are both on the scorecard.

• Design and train state-of-the-art models targeting low-power inference hardware, including SSMs, Transformers, hybrid sequence models, and ML approaches to signal processing problems
• Hardware-software co-design with hardware and compiler teams, so that model architecture, quantization strategy, memory layout, and operator selection are decided together, not sequentially
• Optimize models for deployment through quantization, pruning, distillation, sparsity, and operator-level tuning for Akida and adjacent targets
• Evaluate and select application domains where our hardware provides a differentiated advantage, and translate research findings into product direction
• Stay current with the research frontier and bring back advances from the literature, conferences, and open source relevant to our roadmap
• Present research results to internal technical and business stakeholders and externally at conferences, customer meetings, and industry forums
• Publish selectively where it strengthens our technical position, file patent disclosures, and participate in writing contract and grant proposals

• Neural networks, from first principles through modern architectures
• State Space Models including TENNs (Brain Chip's variant), S4, Mamba, and related efficient sequence models
• Transformer architectures including attention variants and efficient inference techniques
• Signal processing fundamentals including linear systems, sampling, filtering, spectral analysis, and stochastic processes

• Radar signal processing
• Biomedical signal classification
• Agentic systems and retrieval augmented generation
• Robotics and visual odometry

• PhD in Computer Engineering, Computer Science, Electrical Engineering, Applied Mathematics, or a related field with 3 or more years of industry or postdoctoral experience, OR an MS with 6 or more years of equivalent depth
• Strong proficiency in Python and PyTorch, with clean, well-structured code others can build on
• A track record of delivering ML work end-to-end, from problem formulation through trained model to deployed system
• Experience collaborating with hardware, firmware, compiler, and business teams
• Preferred: top conference papers, peer-reviewed publications, or granted patents; direct experience deploying models to NPUs, neuromorphic hardware, or DSPs; experience with quantization-aware training and model compression tool chains