As a Machine Learning Engineer, you will take the lead on optimising the functionality, performance, and algorithmic engineering powering our machine learning models. This is a critical role for a hands-on expert who can deliver state of the art models, set the standard for performance, and ensure our models are accurate, robust, efficient and scalable.
What you’ll do:
- Lead the development and refinement of novel machine learning architectures and algorithms harnessing our nonlinear dynamics. Building deeper network architectures that maximise efficiency and performance.
- Design, build and test models both on device and using in-house simulation framework.
- Collaborate closely with the wider photonics and hardware team to design and evaluate general metrics to assess the computational properties of the hardware and optimise for computational performance.
- Research state-of-the-art machine learning & machine vision techniques and adapt them to be compatible with our hardware.
- Experience:
- Proven track record of developing novel algorithms (papers in NeurIPS, ICML, ICLR, CVPR, or Nature/Science journals)
- Hardware Aware ML / Neuromorphic Computing:
- FPGAs, ASICs, analog computing chips, spiking neural network (hardware), edge AI
- Unconventional training algorithms
- Reservoir computing, self-constrastive learning, forward-forward learning, evolutionary algorithms, equilibrium propagation
- Physics informed neural networks, or applied ML to physics problems
- Deep understanding of mathematics, algebra / topology - key words are latent space, intrinsic dimensionality
- Experience working with hardware as well is a bonus