Applications & use cases

Turn public-sector problems into traceable decisions.

Each use case starts from an operational problem: target, classify, reconstruct, prioritize or justify. The goal is not to add another score, but to make a decision more readable, defensible and useful to teams.

Discuss a concrete use case →
Customs Hub

Concrete mini-cases for customs administrations.

These scenarios show how MAATAI Sentinel, MAATAI Nomen and MAATAI Forensic fit into an existing decision chain without replacing the officer.

MAATAI Sentinel

Justify risk-based selections

ProblemDeclarations are selected without sufficiently readable justification.
DataDeclarations, operators, products, values, origin and control history.
MethodHybrid scoring, business rules, explainable signals and feedback loops.
OutcomeA motivated, traceable selection that officers can use.
MAATAI Nomen

Review multilingual product classification

ProblemProduct descriptions are vague, inconsistent or poorly translated.
DataInvoices, manifests, product labels, HS codes and operator history.
MethodMultilingual analysis, candidate codes and documentary consistency checks.
OutcomeA documented classification note, without automatic decision-making.
MAATAI Forensic

Reconstruct a complex case

ProblemFacts, documents, actors and decisions are dispersed.
DataDeclarations, documents, controls, decisions, overrides and communications.
MethodTimeline, actor graph, audit trail and version preservation.
OutcomeA readable case file for audit, post-clearance control or investigation.
AI governance

Audit an AI-assisted decision

ProblemAn administration must explain what AI recommended and what humans decided.
DataScore, model version, factors, human decision and control outcome.
MethodAudit journal and clear separation between recommendation and decision.
OutcomeClear, defensible responsibility.
Territory Hub

Three territorial mini-cases: VELODIT, ECODIT and FRAÎCHEUR.

The Territory Hub is not limited to VELODIT: it applies the same discipline to three public-decision families — cycling infrastructure, school surroundings and urban cooling. Same foundation: open data, fine grid, priority separated from ease of action, and readable justification.

VELODIT

Prioritize cycling investments

ProblemToo many local requests and too few shared criteria.
MethodNeeds × continuity × safety at 200 m grid-cell level.
Outcome583 sectors cumulating the 3 issues and 17 strategic missing links.
ECODIT

Make school surroundings safer

ProblemChoose which schools to address first without confusing need and convenience.
MethodRoad danger × residents’ needs × lack of calming, with actionability kept separate.
Outcome409 school sites ranked, 40 in the highest relative priority band.
FRAÎCHEUR

Prioritize urban cooling

ProblemDecide where to focus planting, shading, unsealing and cool-place access.
MethodOverheating × vulnerability × exposed population × access to cool resources.
Outcome16,474 grid cells analyzed and 1,579 in very high relative priority.