Research · Engineering · AI

Research &
Development.

We use AI to engineer capabilities and systems that do not exist yet. Research defines the frontier; engineering builds the path towards it.

We do not simply
optimise the present.
We engineer the possible.

We start from a capability that does not exist yet, break it down into testable hypotheses and build models that learn from the domain, simulate alternatives and adapt to the individual or the process.

Our commercial platforms are our working laboratories. MNEMOS is the most radical frontier of this method: AI not as an added feature, but as an instrument for designing what becomes possible.

01 / VERTICAL AI

Models that know the domain

AI adapted to the language, the documents and the constraints of real sectors.

02 / DATA ENGINEERING

Structured and unstructured data

Pipelines that connect databases, documents, events and external sources.

03 / HUMAN-CENTERED AI

Automation under control

Visible sources and a human presence in the decision-making process.

04 / PRODUCT SYSTEMS

From model to product

Inference, observability, security and interfaces for AI in the field.

PROJECT MNEMOS

The memory
that reopens.

Memory Narrative Encoding & Modulation Observational System. Our exploratory programme combines AI and experimental research to study how memory processes can be acted upon, and how the behavioural responses generated by trauma and adverse events can be changed.

The frontier we pursue is the controlled grafting of new memory traces able to sit alongside, transform or counteract the existing ones. Today that capability does not exist in any reliable or clinically usable form: MNEMOS studies how to make it possible and how to address, from the outset, its clinical, ethical, legal and social implications.

Enter the MNEMOS section →

A / THE QUESTION

What we actually know

What is established, what is still contested and what is simply misunderstood, kept apart from one another.

MNEMOS →

B / INTELLIGENCE LAYER

AI inside the process

Individual modelling, adaptive cue selection, probabilistic state estimation, simulation and causal analysis: with traceability and human oversight.

The intelligence layer →

C / DIRECTIONS

Where we can explore

Six testable hypotheses, the translational path, the neural horizon and the ethical boundary, argued rather than assumed.

The directions →

From the laboratory
to the products.

Four research programmes, four independent platforms, a single technology infrastructure.

FINANCE INTELLIGENCE

MurphFin

Modelling of financial and economic data.

LOGISTICS INTELLIGENCE

Trakka

Workflow automation for road transport.

COOPERATIVE TECHNOLOGY

CoopSuite

Purpose-built infrastructure for cooperatives.

INFORMATION INTELLIGENCE

Cryterio

Real-time verification of information.