Sentient Technologies is globally recognized as a leading provider of mission critical technology for lenders, by introducing scalable, instantaneous lending solutions that improve financial outcomes and accelerate origination workflow.

Babak Hodjat is co-founder and chief scientist of Sentient Technologies, responsible for the core technology behind the world’s largest distributed artificial intelligence system. Babak is a serial entrepreneur, having started a number of Silicon Valley companies as main inventor and technologist. We caught up with him ahead of his talk at the Deep Learning Summit this month.

What are the key factors that have enabled recent advancements in deep learning?

Much of the fundamental technology has been around for quite some time. What is different now, however, is that the computing power has dramatically increased while the cost to scale data has dramatically decreased. These economies of scale in computing have made it much more possible for us to see smaller organizations attack more complex problems in a more complex manner.

Adding to the reduced barriers in computing scale, within this past year alone we’ve seen significant funding and the rise of new startups that are creating many specialized solutions to a variety of industries. What we’ve observed is that deep learning is still only a single tool that, on its own, is not sufficient to solve challenging decision-making problems.

While deep learning will allow us to use many kinds of unstructured data including images, our goals are fundamentally larger. We are building intelligent decision-making systems, which is different than what others are doing.

What are the main types of problems now being addressed in the Neural Network space?

Neural Networks, and Deep Learning in particular, are being widely applied to image classification and feature extraction problems. The applications have not remained exclusively in the image-processing domain, however, and text and speech based applications have also been reported (e.g., improvements to speech recognition systems). We believe that Deep Learning systems are a good complement to other AI methods such as Evolutionary Algorithms, providing information rich features for EAs to construct decision making rules.

What are the practical applications of your work and what sectors are most likely to be affected?

While our DAI has very broad applicability, we have been laser focused on its application in trading, and only recently have we started exploring, through a research collaboration with MIT, its applicability in other areas such as healthcare (e.g., blood pressure prediction from raw readings in an ICU setting).

What advancements excite you most in the field?

We’re most excited by both the massively distributed artificial intelligence as well as the more intelligent decision-making that will come with the technology. We’re increasingly faced with a rising number of decisions that we must make on a daily basis – more so than past generations – and with recent breakthroughs in AI we’re able to focus much more on the decisions that matter. This can only be enabled through massively distributed computing. We can not only produce very complex analysis, we can also validate them and affect them in high dimensional problems using massively distributed compute capacity.

We’re pioneers in applying the “compute” of big data. It’s exciting that we’re at the cusp of solving some very interesting problems that we’ve never been able to before. We’re starting to realize that every problem is an artificial intelligence problem, and we’re only just beginning to see the scale and scope of their practical applications.

The Deep Learning Summit is taking place in San Francisco on 29-30 January. You can get more information and register here.