Public affairs
Geographic, industry,
and microcultural segmentation
Detects demands that are
not yet on the agenda.
A demand rarely starts on the agenda. It starts in one district, around one concrete problem, among people who have something in common, and it has been growing for some time when it reaches the general media.
An average of the whole conversation hides it. The platform sorts what is published openly along three axes, where it happens, what it is about, and who is speaking, so that an institution or a company sees a demand while it is still local and still small.
Three axes, one map
Every signal the platform captures is placed on three axes at once. The axes are defined with the organization at the start.
Geographic
Where it happens. The reading goes from the territory as a whole down to the municipality and the district.
Industry
What it is about. For an institution, the matters that concern the public; for a company, those that affect its activity.
Microcultural
Who is speaking. Communities that share a trade, a pastime, or a way of life, and that age or income would not describe.
Sources the platform reads
Media, social, search engines, AI assistants, and proprietary data.
60 min
shortest data refresh cycle
What an institution or a company gets
Segment map
Each crossing of territory, topic, and community is a segment. For each one the platform shows how much is said, in what tone, and how it is changing.
Segments nobody defined
Automatic clustering groups conversations that resemble one another and brings up segments that no one had foreseen.
Topics by territory
For each territory the platform shows which topics are being discussed, which are emerging, and which are fading.
Emotional climate
The emotional spectrum is read by topic and by audience. The same topic can carry concern in one segment and indifference in another.
Artificial activity
It identifies bots and coordinated account networks. Volume manufactured by a handful of accounts is not a public demand.
Social studies
For a public body that commissions a social study, the same map describes what is said openly about a matter, by territory and by community, with every fact traced to its source. It describes a conversation; it does not measure a population.
How the platform works
The platform works in a cycle of seven steps: listen, normalize, understand, anticipate, decide, act, and learn. Segmentation depends on the second: every source is mapped to one common schema before it is analyzed.
Six specialized AI agents analyze the same signal. One detects weak signals and volume anomalies, which is how a demand is seen early. Another identifies bots and coordinated account networks, so that manufactured volume is not taken for a demand.
The six reports are cross-checked before they reach a person.
Which demand to attend to,
and where.
With that reading, the institution or the company decides which demands to attend to, in which territory, and which to keep following without doing anything yet.
Not every signal calls for action, and not every action should be public. Each signal ends in one of three outcomes, based on thresholds.
Silence
Noise is tagged and archived. This is the default.
Internal alert
With context, history, and a recommended response.
Public action
Always previewed before it is published.
How an AI assistant
describes you.
First impressions no longer come only from a search engine. They also come from an assistant. Anyone asking about an institution or a company, or about a matter that affects it, already gets an answer written by an AI. The organization needs to know what it says, and which sources it draws on.
The platform audits every search result and every assistant answer. Where information is missing, the gap is filled with original sources, which the organization publishes in its own name, through its own channels, and after reviewing them.
Security, traceability,
and human decision
A map of demands shows where an institution or a company stands with the public, so access to it is restricted: by invitation only, with MFA and enterprise SSO, and with six roles with granular permissions by section. Every action and every AI query is logged, and data is encrypted in transit and at rest.
GDPR and the EU AI Act ask where each conclusion comes from and who makes the call. Here, that is built in from the start.
Every fact has a source.
The platform proposes.
A person decides.
Nothing ships unreviewed.
AI amplifies human judgment. It does not replace it.
In production in six months.
Typical schedule. Two-week sprints, a demo every two weeks, and incremental deliveries you can verify.
Months 1–2
Foundations
Discovery, architecture, corporate access, and the first modules in use.
Months 3–4
Measurement
Analysis modules, the first version of the AI, and the executive dashboard.
Months 5–6
Optimization
Early warnings, user testing, training, and launch.
Afterward
Evolution
New modules, API integrations, and handover to full autonomy.
Frequently asked questions
What are the three axes of the segmentation?
Where it happens, what it is about, and who is speaking: territory, topic, and community. Every signal is placed on the three at once.
What is a microcultural segment?
A community that shares a trade, a pastime, or a way of life, and that age or income would not describe. Some are defined with the organization at the start; others come up through automatic clustering.
How does a demand that is not yet on the agenda show up?
As a weak signal: a conversation that grows in one segment before it reaches the general media. The platform detects weak signals and volume anomalies, and sets apart the volume manufactured by coordinated accounts.
Is it useful for a social study?
Yes. For a public body, the map describes what is said openly about a matter, by territory and by community. It describes a conversation; it does not measure a population.
Does the platform publish or respond on its own?
No. The platform proposes and a person decides; nothing ships unreviewed, and silence is the default. Whatever is published, the institution or the company publishes, in its own name, through its own channels, and after reviewing it.
How long does it take to be up and running?
The typical schedule is six months, with the first modules in use in months 1 and 2. Early warnings arrive in months 5 and 6, after the analysis modules.
We do not name our clients.
Confidentiality is part of the service: here we speak of industries.