I have done industrial strength traditional AI for 20+ years: I wrote three expert systems for product configuration verification for cisco, and did NLP (Natural Language Processing) for Kanisa, Google, Kosmix, and Sensai. I also have worked with programming language definition and implementation for Dialogic Systems, Kanisa, and Sensai. I know five natural languages.

My main research interest since 2001 is a kind of next-generation Artificial Intelligence systems called "Understanding Machines". They are based on an algorithm called "Artificial Intuition" that provides Understanding - the base skill in language understanding, motor control, vision processing, etc. This should be contrasted with Reasoning, which is a logic-based skill that requires Understanding as a base - it anchors all the variable we want to reason about. Classical AI research started with the misguided notion that Reasoning was all there was, since that's the part we can see using Introspection. But Understanding actually consumes over 99.9% of the brain's cycles and is the real basis of intelligence.

Understanding is instantaneous, precedes Reasoning in the evolutionary record and precedes Reasoning temporally in problem solving; you cannot reason about that which you don't yet understand. Understanding is invisible to introspection and thus subconscious, it is instantaneous and parallel. Reasoning is the only kinds of thinking we can directly observe but it is a slow, step-at-a-time process that is used only in planning and in a few other classes of situations. For more, see my main link in my user info or read Daniel Kahneman's book "Thinking Fast And Slow" whcih also discusses this "Dual Process Theory".

Since 2013 I'm the CTO and co-founder of Sensai Corporation ( http://sens.ai ) where I have designed and co-implemented a Data Science Platform for unstructured content in the form of text. The system has a Java backend connected to MongoDB and a JavaScript/jQuery/SlickGrid front end.