brain cutout.png


Our software implements a remarkable idea from cognitive science – that thought is a combinatorial system.


Language is serialised conceptual structure. Conceptual structure encodes everything we know about the world.


Conceptual structure is a universal data structure. It can standardise the world’s data. It can enable true Artificial General Intelligence.


What do we do?

High Level Architecture.png

We’re an early-stage startup with non-fiction answers to science-fiction questions.

Can computers really represent knowledge? What about common sense? Will they ever understand English like we do? What about Chinese? Could they translate between languages as accurately as people can? And how are vision, language, and movement connected? Why can we use English to teach people how to do things but not robots? Can a computer know why it did something or how it knows something? What exactly is data? And why don’t voice assistants work like in the movies?

At the heart of the answers to these questions sits an undiscovered technological linchpin: conceptual structure. This is our secret sauce. We’ve used it to build a bot that can respond to questions in English about what it sees in a virtual 3D environment. We’re extending its abilities so it can follow instructions in English, too. Our long-term mission is to realise the awesome potential of this technology, which we believe will cause a paradigm shift in computing equal to the invention of the internet, culminating in a true Artificial General Intelligence. 

But to get there we need to take the first step. Our short-term goal is to commercialise this technology in a relatively small, uncontested market that allows us to do some societal good: reducing admin costs in the NHS. We are working with a development client to build a reporting solution that automates the reading of clinical notes, providing valuable answers at a fraction of the cost. The lessons we learn there will allow us to automate the process of reading almost anywhere else, from financial reports to legal documents to Wikipedia.

Who are we?


Des Kelleher

Founder and MD, Des spent five years researching Cognitive Science/Linguistics and developing a system to represent knowledge and understand language. His previous company built business intelligence software for the NHS and was acquired by a FTSE 100 company. He has a background in Computer Science and Mathematical Physics.


Jayden Ziegler

Jayden completed his Ph.D at Harvard University with research focused on the structure, origins and development of language. He brings world-class expertise on conceptual structure and word meanings to bear on Mental Technologies' real-world implementation of this research. Jayden previously studied at NYU and Princeton University.


Christoffer Olsen

Chris was employee #5 at Kalo, a London-based startup that raised $20m based on a SaaS application that he built and oversaw. An expert software developer and product manager (as well as a former politician in his native Norway) Chris brings broad-ranging technical, business and people skills to Mental Technologies.


Mike Stukalov

Mike has a master’s degree in Cognitive Science and years of experience implementing ML and deep-learning systems in the financial sector. He advises Mental Technologies on project management, the existing state-of-the-art and where our technology can fill gaps in the market.