Oliver Grainge

AI Research & Engineering

I’m Oliver Grainge, a researcher and engineer specializing in efficient machine intelligence. I develop practical AI solutions that work in the real world—focusing on neural network compression, quantization, and optimization techniques that make deep learning accessible on resource-constrained devices.

Current Research Focus

My work centers on making AI more efficient without sacrificing performance:

Ternary & Low-Bit Quantization — Pioneering training methods for ternary (−1, 0, +1) neural networks that achieve state-of-the-art accuracy while dramatically reducing computational requirements for edge deployment.

Neural Network Compression — Developing structured pruning and quantization strategies specifically optimized for visual place recognition and computer vision applications.

Hardware-Aware Optimization — Creating deployment-ready solutions that bridge the gap between academic research and commercial applications on embedded platforms.

My research has been published in venues like IEEE Robotics and Automation Letters and implemented through partnerships with Foster + Partners and Arm.

What You’ll Find Here

Research & Publications — Deep dives into my latest findings in neural network optimization, complete with reproducible code and detailed technical discussions.

Open Source Projects — Production-ready tools and libraries that make efficient AI techniques accessible to researchers and practitioners.

Technical Writing — Insights from the intersection of research and real-world deployment, covering everything from quantization theory to edge AI implementation.

Collaboration & Contact

I’m actively seeking collaborations on open-source AI projects and research that pushes the boundaries of efficient machine learning. Whether you’re working on edge AI, neural compression, or related challenges, I’d love to explore how we might work together.

Building the future of efficient AI, one optimized model at a time.