Today we published our first paper.

It presents the knowledge representation model we have been developing and implementing for many years, and formalizes ideas that have been part of our work for more than two decades.

This is also the first time we are making the model public in this form. We have kept much of this work relatively private for a long time, so publishing it is a significant milestone for us.

And, on a more personal note, this is my first paper. Not bad for 50.

I also want to recognize Olga Valls, co-author of the paper, and all the engineers who have worked with us over the years and made the technology and its implementation possible.

The paper covers the knowledge model behind most of our products and solutions.

It is the first of four papers. The next one will address the NLP side, followed by a complete description of the cognitive architecture, and finally how we are exploring a hybridization with a local, small LLM to produce elegant verbalizations and natural conversations with a machine that knows and understands the knowledge stored locally on your computer, without requiring a server.

The paper is available here. Feel free to read it, download it, and let us know what you think.