If the answer to your question is not listed here, just contact us and we will give you all the information you need.


How much do we cost?

We believe in value-based consulting. We seek to quantify cost savings, revenue growth, and productivity gains. Focusing on the value that we add to your business ensures that you get value for money. We may well be able to help you expand your business in ways that you have only dreamed of in the past.


Do we operate internationally?

Yes. Through the API group we have offices in a number of countries and provide you with the support that you need in the country in which you are based. Places where members of our team have carried out consulting work in the past includes; the USA, Israel and Japan.


Who are we?

API Cognition is a small start-up specializing in software for the simulation of artificial cognition. We are affiliated with the API group of companies, which includes; API Engineering, PPI Engineering, API Capacitors, CNE International and API Stone. International and API Stone.


What have we got?

Fundamental artificial intelligence research addresses problems of recognition or regression. More advanced techniques integrate reinforcement learning algorithms to solve 'action selection' problems. However, such systems are not capable of 'understanding' in the sense that we commonly recognize it. API Cognition has a solution to this problem; a model that subsumes all of these
approaches (whilst additionally incorporating logical inference) within a single coherent framework.


How?

We have conducted privately funded research dating back almost 20 years. During this period we have worked on; robotics simulation, neuroscience, and artificial neural networks. We employ an iterative
approach whereby we; (a) seek to identify simple principles of brain function, (b) develop models and algorithms that are consistent with neuroscience research, and (c) implement them in an embodied agent. This work has culminated in a 'whole-brain' model incorporating all major brain regions.


What software do we use?

We have built specialized libraries that fall into one of two categories; novel neural networks and algorithms, and secondly, generic software components for running complex neural network and robot simulations in a multi-threaded plug-in architecture.


What is our engineering approach?

For testing and simulation we use a multi-threaded plug-in framework comprising loosely-coupled components. This has three benefits; it facilitates design conception, simplifies testing (as components can be tested independently), and allows for a division of engineering resources (as components can be developed in parallel). The brain is less amenable to this kind of approach since neurological systems are massively interconnected. For this we need a complementary strategy. Firstly we establish simple principles of brain organization that relate to the brain as a whole. This allows us to understand how the brain works, at a meta-level, rather than at the component level. Secondly, we test the brain model through building
different combinations of functional groups. This allows us verify that each functional group is operating as intended. Thirdly, we are able to assign engineering resources to libraries that relate to differing levels of abstraction within the system, rather than to individual components.


Patents?

The complexity of our whole-brain model presents challenges when it comes to writing and filing patents. Fortunately our engineering approach allows us to identify methods that correspond to key cognitive building blocks. We are then able to devise hardware and systems that incorporate these methods. Provisional patent applications were filed in 2016, non-provisional patent applications were filed in 2017.


Commercial exploitation?

We do not believe in research for the sake of research itself. In order to avoid the pitfalls of developing in a vacuum, absent of commercial considerations, we have been active in identifying markets in which this technology can be most readily exploited and where existing user needs are currently unfulfilled. We have in particular (but not exclusively) focused on the health sector. We do however have a range of expertise readily accessible for entering significantly different markets. This has enabled us to investigate the potential of this technology in; autonomous robotics, the video games and animation industries, and hardware projects relating to human-user interaction.