Where Deep Learning Ends Deep Thinking Begins

Screening the Human Machine

Oxford Immune Algorithmics (OIA) is a biotech company that has built one of the most impressive AI technology stacks to decipher one of the most sophisticated adaptive natural systems, the immune system. This is to help people track, understand, and take decisions based on their personalised data. The ultimate goal is to enable early diagnosis of any (preventable) disease or condition and avoid the need for expensive and often deadly post-diagnostic treatments by actively collecting and analysing data, and exploiting the superhuman capabilities of the kind of AI-relevant technology that we are creating  for use in medical applications.

Research-driven Deep Medicine

The keys to understanding and dealing with complexity are monitoring, deep understanding, and simulation. We have designed state-of-the-art methods to reprogram systems based on solid, cutting-edge  research. We have designed and implemented novel techniques for fast pre- and post-processing of data to track essential information and provide decision-making tools for personalised medicine. Our techniques involve algorithmic machine learning, model-driven discovery, computer simulation, complex network systems, and computational medicine, bridging research, science and industry.

With about two thousand members, our AI meetup groups in the UK, are among the largest in their cities, and it demonstrates our engagement to connect the academic and industrial communities:

London

Reading

Oxford

Mind + Machine vs Disease

We aim to connect academia and industry not only by introducing traditional tools from machine and deep learning into healthcare in order to go beyond what is currently possible in the practice of medicine, but also by finding ways to match biological markers and diseases using approaches that surpass the current capabilities of statistical machine learning--which is unable to perform algorithmic inference--and thereby generating cause-and-effect models of human diseases. Our pioneering team, with its solid research and strong publication record, is cracking healthcare.

Our technological pipeline involves performing massive in silico simulations that allow us to build a constant stream of artificial data that is tested against empirical observations to produce candidate generative models, which in turn would be tested in a continuous model-data driven cycle. What is learned from both populations and individual cases then serves as knowledge which constitutes the basis of model-driven solutions for detection and prediction.