News and Announcements
Interactive Engineering makes its semantic technology self-extensible using human-readable resources
- Published January 14, 2020 12:00AM UTC
- Publisher Wholesale Investor
- Categories Company Updates
Interactive Engineering is building its vision of an active, undirected structure in a computer to allow a machine to read and understand text, while expanding what it can understand by using human-readable resources with a full meaning.
Another strand is to make subsystems reconfigurable through words, in the way you do what the sign says, without knowing how you do it.
Interactive Engineering Investment Highlights
In 2014, we built a system for handling queries for the US Health Insurance industry – they field a billion calls a year, each company has tens of thousands of different policies, each slightly or completely different.
We built that first system by hand and secured a licensee in the States, who beat off the local competition to get a contract with one of the largest health Insurance companies.
Problem
Words carry meaning, sometimes lots of meanings. AI hasn’t attempted to handle this problem, instead running on associations or statistics or key words – great when nothing changes, like the rules of Chess, or the tasks are simple, like turning up the volume, but potentially deadly when tasks are complex, and a single misstep could presage disaster, or things change rapidly, which is where we seem to be headed.
Market Size
We will be offering a product which knows the meaning of words and can extend its vocabulary by using a human-readable dictionary.
It is not aimed at a niche, or to compete with existing products – there aren’t any.
Competitive Advantage
We use a machine to build Active Structure – a structure which can modify itself, extend its knowledge using human-readable resources, and run analysis on “unprecedented” things. We are interested in new or rapidly evolving environments – you don’t need to go to Mars to find one, you are living in one.
People have a very low limit for handling new information – Active Structure will allow us to break that limit.
Active Structure can also handle intellectual drudgery better than people can, like ploughing through a million Google hits.
Revenue Model
The product is general purpose, which means we need to find licensees in many different areas, and train them in the technology. They probably won’t have one employee we would allow to work on our stuff – we are talking about changing from programming of simple messages or data to a machine “understanding” and handling complex messaging.
A licence will include a person who can guide in the hiring of appropriate workers, their training, and the matching of the system’s abilities to applications.
Team
Our director James Brander has thirty years of experience with Active Structure. The team also includes a programming expert in Delphi, a solutions architect and a software engineer. They have been selected for their ability and interest to work with new mind-stretching concepts.
Exit Strategy
We expect to create a public company to manage the licensees, with the R&D company staying private.