03 · Platform
Three platforms, one value chain
REFLEX, XAORUS and XIGMA, designing a solution, acting on what it finds, and improving the models behind it.
REFLEX · XAORUS · XIGMA
Three proprietary platforms, each addressing a different layer of the same problem: designing a solution, acting on what it finds, and improving the models behind it.
01 · Design and deployment
REFLEX
A no-code and low-code environment to design, orchestrate and deploy AI detection and recognition solutions.
REFLEX is where a solution is defined. It exists because the person who knows what needs detecting on a site is rarely the person who would otherwise have to write the software to detect it, and the distance between those two people is where most industrial AI projects are lost. Bringing design, orchestration and deployment into one environment shortens it.
- 01
Design of detection and recognition solutions without writing the implementation.
- 02
Orchestration of how those detections are composed into something operationally meaningful.
- 03
Deployment of the result into the target environment.
- 04
The same environment for all three, so design and what is running do not drift apart.
02 · Events, alerts and incidents
XAORUS
An event, alert and incident-intelligence platform enabling real-time decision flows, auditability and multi-channel notification.
XAORUS is where a finding becomes something someone acts on. Detection is the easier half of the problem; the harder half is that a finding has to reach the right person, through a channel they actually use, quickly enough to matter, and leave a record that holds up when the incident is reviewed weeks later.
- 01
Event, alert and incident intelligence, rather than a raw detection feed.
- 02
Real-time decision flows, so a finding leads somewhere rather than accumulating.
- 03
Auditability, because an incident review asks what was known and when.
- 04
Multi-channel notification, reaching people where they already are.
03 · Training and specialisation
XIGMA
A deep neural learning core for advanced training, re-training and specialisation of models.
XIGMA is the reason a model can be specialised to one environment rather than configured for it generically. It is also what makes independence operational rather than rhetorical: training and re-training held in-house are what allow a model to be owned, and to be re-specialised when the environment it watches changes, as a line, a yard or a season inevitably will.
- 01
Advanced training of models for a specific operating environment.
- 02
Re-training, as the environment the model watches changes over time.
- 03
Specialisation, so a model fits one site rather than approximating many.
- 04
An in-house lifecycle, which is what ownership of a model actually consists of.
Each description below expands the platform's own published definition and goes no further. There is no architecture here, no topology, no data flow between the three, because a product description is not an architecture, and inferring one from three product names is how a public page ends up asserting something nobody verified.
Related environments
Related topics
Discuss a deployment
Which platform is relevant depends on the environment and the outcome. That is a conversation rather than a configuration.
Industrial and enterprise enquiries · EN / ES