MEET OUR 2018 CODE_n CONTEST FINALISTS: thingsTHINKING from Germany
What do humans do better than computers? Think with common sense. Capturing the meaning of data is the focus of thingsTHINKING technology. CEO Dr. Sven J. Körner tells us how they teach their machines to “understand” the meaning of text using semantic processing to interpret data to deliver real knowledge.
Jessica: What is thingsTHINKING all about?
Sven: Common sense. For computers.
Humans are way better than computers when it comes to common sense and understanding natural language. thingsTHINKING brings common sense to computers. thingsTHINKING understands relevant and useful information in your unstructured data, not unlike a human. This accelerates finding things, making contextual decisions, and – as we humans like to say – doing the job right. Think of us as a sidekick, a forklift for your brain.
Jessica: How did you come up with the idea?
Sven: The idea grew out of a project at Karlsruhe Institute of Technology (KIT) where we researched the subject for over a decade. We started in a narrow field and then realized that semantics and the processing of meaning is applicable to pretty much anything we humans do every day, especially where unstructured information is being processed. Well … and here we are.
Jessica: What challenge or problem are you trying to solve?
Sven: We have too much unstructured data. And often the information within that data is not implicit, i.e. not stated in the document/text itself. Instead, it’s interpreting the data that delivers the real knowledge. That used to be a job reserved for humans. The problem now: too much data, not enough humans!
Our semantic processing helps customers with their accounting, tax, M&A, and due diligence processes, or in requirements engineering or contract processing, or sometimes in areas we didn’t even know this would be possible.
Jessica: In which sectors of industry can your solutions be used, and in what way is your technology innovative?
Sven: Semantic processing seems to be relevant to lots of areas. This broad applicability is also reflected in our customer base, which ranges from automotive companies to the Big Four accounting firms, but also law firms and insurance companies. A major human trait we try to teach to our machine is to “understand” the meaning of text but throw away the words. Obviously humans are already better at that, but our solution can assist humans – that’s why we describe ourselves as a forklift for your brain.
Also, our system comes pre-trained, which is the biggest difference compared to “classic” AI approaches. If you want to use thingsTHINKING, our aim is for the first version to be up and running for you within 15 to 20 days. Get your hands dirty quickly and deliver results; don’t waste time with PowerPoint presentations and meetings … get things done, find out what works and what doesn’t. This a) helps the customer understand what you do and b) it addresses any fears and objections they have to AI.
Jessica: Thank you for this interview, Sven!
thingsTHINKING (Semantic) Forensic Investigation
thingsTHINKING (Semantic) Legal Tech Example