Jobs at Intellegens

Intellegens, a spin-out from Cambridge University, has developed a unique Artificial Intelligence (AI) method for training neural networks and generating predictive models from incomplete data. This new approach overcomes a common problem in current deep learning that requires all the data to be available to train accurate models. Typically, incomplete data needs to be cleaned or discarded, resulting loss of valuable knowledge. The fragmented data problem is a cross market, widespread problem. You can think of all experiment based technologies as being sparse data problems ie you do experiments to find more data, we do this via the models. Some typical projects include: - design of super alloys: where the technology was first developed, historical material data sets are heavily fragmented but still contain valuable information. Generating models from the data allowed materials to be designed to achieve specific operating properties - drug discovery: molecule / protein activity is a very sparse data space where drugs are only tested against a small set of proteins, however there is a lot of available data to learn from. There are numerous other opportunities within pharma / experimental space. - health care: patent data is naturally sparse, with each individual having a unique profile with varying numbers of data points we can train models to predict all outputs from varying sets of inputs. We are also looking at problems in predictive maintenance scheduling, autonomous vehicles and insurance to name but a few. We have a number of exciting ways we are looking to extend the core technology around noisy data and time series data as well as how the technology can be applied cross domains at the data layer, with a brand new approach to the federated learning approach. This is a very exciting opportunity to be an early employee at an up and coming tech company with real science, real customers and real potential.