KAT-AI is the Artificial Intelligence engine embedded in every ZEP-X terminal by RETICEO. Running entirely on the RETICE local network — without Internet, without external cloud — KAT-AI operates across three fundamental dimensions of school life: Adaptive Pedagogy, Intelligent Governance and Learner Guidance.
KAT-AI is an Artificial Intelligence system locally embedded in every ZEP-X terminal and in the Smart Local Cloud of the RETICE infrastructure. It depends on no external cloud service, no telecom operator and no Internet connection to function.
Developed by RETICEO (subsidiary of KA Technologie), KAT-AI leverages data generated locally within the school — learning behaviours, KAT-EXAM assessment results, attendance records, progression on the KA SOFT LMS — to produce high-value analyses, recommendations and automations.
Its three modules — Pedagogy, Governance and Guidance — cover the entire school lifecycle, from the first day of class through to professional integration.
KAT-AI Architecture — Local Processing
Three complementary modules that transform every school into an intelligent, autonomous ecosystem centred on the success of each learner.
KAT-AI revolutionises teaching practice by analysing in real time the learning data of each student to deliver an individualised, adaptive and measurable teaching experience. The teacher is no longer alone in facing classroom heterogeneity — KAT-AI assists at every moment, from lesson preparation to automated assessment correction.
KAT-AI analyses each student's data to build individualised learning pathways adapted to their needs, pace and learning style. Each learner receives exactly the content they need, at the right moment.
KAT-AI dynamically adjusts the difficulty level of lessons based on observed progress. If a student masters a concept, the system advances. If they struggle, it offers reformulations, alternative examples and consolidation micro-exercises.
KAT-AI exploits all data generated on the KA SOFT LMS and in KAT-EXAM to identify areas where students need support and provide actionable insights to teachers and management.
KAT-AI handles time-consuming tasks that divert teachers from their core mission: automatic grading, homework management, exercise generation and progress report creation.
KAT-AI acts as a virtual tutor available at any moment for each student. It answers questions, explains errors, suggests problem-solving methods and encourages progress — without Internet connection.
By analysing behavioural patterns (absences, connection time, scores, engagement), KAT-AI early identifies students at risk of dropping out and alerts teachers and management before the situation becomes critical.
Real-world scenario — Adaptive Pedagogy
KAT-AI transforms a standard teaching sequence into a personalised experience for each student:
The teacher starts the lesson on fractions. KAT-AI analyses each student's historical data — previous homework results, recurring errors, completion time — and automatically generates 3 levels of sequences: consolidation, standard and advanced.
During exercises, KAT-AI detects that 7 students are struggling with the same concept (fraction division). It alerts the teacher on their ZEP-X dashboard and simultaneously offers these 7 students an alternative explanation with visual examples.
At the end of the class, KAT-AI generates a class report in 30 seconds: skill mastery rate, list of students to review, content suggestion for the next session, and differentiated homework sets generated automatically by level.
The next day, each student finds on their ZEP-X terminal personalised exercises aligned with their identified gaps. 2 students whose risk score has risen are flagged to the educational counsellor for an individual meeting.
KAT-AI equips school principals, academic inspectors and ministries with a real-time intelligent management system. No more decisions made blindly on outdated data — each stakeholder has a personalised dashboard, automatic alerts and AI-assisted planning tools.
KAT-AI generates contextual dashboards for each governance level: school director, head teacher, academic inspector, ministry. Each stakeholder sees exactly the indicators they need, in real time.
KAT-AI automates low-value administrative tasks: timetable management, attendance tracking, official document generation, enrolment and student record management — freeing the administrative team for higher-value activities.
The AI engine continuously monitors key indicators and triggers intelligent alerts when a situation requires a manager's attention — before the problem worsens.
KAT-AI goes beyond reporting — it predicts future trends by leveraging the institution's historical data. A principal can anticipate next year's exam results as early as January and adjust their pedagogical strategy accordingly.
KAT-AI optimises human resource allocation in the institution: identifying teacher training needs, monitoring pedagogical performance, planning cover and detecting workload overloads.
KAT-AI aggregates data from all institutions connected to the RENAL-SMART network to provide ministries with a consolidated, real-time view of the national education system — without intermediary, without delay, without data distortion.
Real-world scenario — Intelligent Governance
A school management day assisted by KAT-AI:
On arrival, the principal opens their KAT-AI dashboard on their ZEP-X terminal. They receive 3 automatic alerts: 34% absence rate in Year 13 Science (vs 12% normally), declining results in Physics-Chemistry (−18 points over 2 weeks), a teacher absent and unsubstituted for 3 days in Year 10.
KAT-AI suggests immediate actions: assign the duty teacher to Year 10, automatically send a notification to parents of absent Year 13 students, and activate the intensive KA SOFT revision module for Year 13 Physics-Chemistry classes.
At 10am, the principal receives the KAT-AI forecast for this year's national exams: 73% estimated pass rate (vs 68% last year), with a list of the 12 most vulnerable students and a personalised support plan for each — ready to activate with one click.
On Friday, KAT-AI generates the weekly regulatory report for the academic inspection — attendance data, results, incidents — with all digital signatures, in 2 minutes, ready to be transmitted via the RENAL-SMART network.
Academic and career guidance is one of the most structuring decisions in a young person's life — and one of the most poorly supported in resource-constrained education systems. KAT-AI bridges this gap by offering each learner a personalised, data-driven guidance journey connected to local, regional and national labour market realities.
KAT-AI builds a rich, evolving profile for each student — academic skills, aptitudes, interests, learning style, behaviours, ambitions — which underpins all guidance recommendations.
Based on the student's profile and historical success data, KAT-AI recommends the most suitable tracks for their skills and aspirations — with a calculated success probability for each option.
KAT-AI integrates a database of careers, training options and job prospects aligned with local and regional labour market realities. Each guidance recommendation is contextualised within the territory's economic needs.
KAT-AI detects signs of stress, anxiety or declining motivation in learners and alerts guidance counsellors and teachers. It also suggests appropriate support resources for each situation.
KAT-AI prepares and assists guidance interviews between the student, their parents and the counsellor. It automatically generates a personalised summary sheet with recommendations, options considered and key questions to address.
KAT-AI tracks alumni in their higher education and professional pathways, enabling continuous improvement of guidance recommendation quality based on real observed outcomes.
Real-world scenario — Personalised Guidance
KAT-AI supports Aminata in her A-level and higher education guidance decision:
KAT-AI has built Aminata's profile over 2 years of schooling on ZEP-X: excellent in Maths and Physics (85/100), passionate about design activities (digital projects on KA SOFT), visual and hands-on learning style, strong resilience when facing challenges. Compatibility scores: Science track (92%) > Technology track (78%) > Economics track (41%).
KAT-AI generates Aminata's "My Future" sheet: 3 recommended tracks (Year 13 Science, MPSI preparatory class, Electronics HND), 5 matching careers (Telecom Engineer, Software Developer, Researcher-Teacher, Network Architect, Data Scientist), with job prospects and average salaries in her region.
During the guidance interview, the counsellor has a complete summary sheet generated by KAT-AI. They can present Aminata and her parents with a clear visualisation of her potential pathway, the entry requirements for her preferred tracks and the actions to take now (Chemistry catch-up, NETSCP projects to showcase).
3 years later, Aminata, in her first year of preparatory class in Abidjan, is added to the alumni base. Her success strengthens KAT-AI's guidance algorithms for future students with a similar profile — a virtuous cycle of continuous improvement.
KAT-AI is not installed separately — it is natively integrated in every ZEP-X terminal and in the Smart Local Cloud of the RETICE infrastructure. The AI is therefore available for every student, teacher and manager, from the moment the terminal is switched on, on the institution's local network.
KAT-AI's software architecture rests on a stack of three integrated layers: the local AI engine (embedded inference), the KA SOFT LMS suite (pedagogical data) and the KAT-EXAM module (assessment data). These three layers share a unified database hosted on the institution's Smart Local Cloud.
ZEP-X Software Stack
All data remains on the institution's Smart Local Cloud — AES-256 encryption, zero external synchronisation
KAT-AI's founding principle is the absolute protection of educational data. In a world where digital giants monetise school data, RETICEO makes the radical choice of local AI — with no external transmission whatsoever.
KAT-AI's AI models run exclusively on the institution's Smart Local Cloud. No data is sent to an external server, public cloud or third-party provider. Inference takes place locally, in real time.
All personal data (student profiles, results, AI analyses, guidance recommendations) is encrypted with AES-256 at rest and in transit on the RETICE local network. Full compliance with local regulations on the protection of school data.
KAT-AI never uses learner data for commercial, advertising or external profiling purposes. The analyses produced serve exclusively to improve learning outcomes and institution governance.
In the RENAL-SMART 80/20 architecture, consolidated (anonymised) data is aggregated at the level of the national sovereign Datacenter (Tier 1) — under the exclusive control of the ministry and the State, with no foreign party involved.
Each learner (and their legal representatives) has the right to access, rectify and delete their AI profile. A dedicated dashboard enables them to consult the data collected and the recommendations generated.
All decisions made by KAT-AI (guidance recommendations, alerts, pedagogical adaptations) are timestamped and logged, and can be audited by management, the academic inspection or a supervisory authority.
| Criterion | Without KAT-AI | With KAT-AI (ZEP-X) |
|---|---|---|
| Individual student tracking | ✗ Quarterly paper report, partial view | ✓ Real-time profile, daily analysis of each learner |
| Pedagogical adaptation | ✗ Single lesson for the whole class (heterogeneity ignored) | ✓ Individualised pathways, level adjusted in real time |
| Assessment correction | ✗ Manual, 2–5 days, no detailed feedback | ✓ Automatic, instant, with explanation of each error |
| Dropout detection | ✗ Noticed after the drop — often too late | ✓ Early, 3–4 weeks before critical drop |
| Management dashboards | ✗ Paper reports, data delayed by several weeks | ✓ Real-time dashboard, automatic alerts, forecasts |
| Academic guidance | ✗ Single interview, subjective, no structured data | ✓ Multidimensional profile, data-driven recommendations |
| Regulatory reports | ✗ Manual entry, several days, risk of errors | ✓ Auto-generated in 2 minutes, digitally signed |
| Data sovereignty | ✗ Dependent on external providers (Google, Microsoft…) | ✓ 100% local, AES-256 encryption, zero external cloud |
| Internet connection required | Variable (often required) | ✓ None — runs entirely on the RETICE local network |
| Additional cost | — | ✓ Included in the ZEP-X / BOO RETICEO offer — zero additional cost |
KAT-AI activates from the very first ZEP-X terminal deployed. Its effectiveness improves exponentially as the RENAL-SMART network expands.
ZEP-X terminals deployed in 1 institution. KAT-AI collects its first pedagogical data, configures learner profiles and activates the 3 modules.
Extension to 10–50 institutions on the RENAL-SMART network. KAT-AI algorithms are enriched by cross-institution data and inter-school benchmarks become available.
KAT-AI feeds the national sovereign Datacenter with anonymised data. The ministry has a consolidated national dashboard to steer educational policy.
KAT-AI becomes the engine of national educational AI — country-level success predictions, data-driven national guidance policy, export with zero foreign dependency.
Zero-Paper terminals in which KAT-AI is embedded — 8 PC + Tablet models, locally assembled by NETSCP, powered by CeRSER.
Discover all 8 ZEP-X models →Digital examination platform — performance data generated by KAT-EXAM feeds directly into KAT-AI's pedagogy and guidance modules.
View KAT-EXAM →The network infrastructure that carries KAT-AI — RETICE intra-school local network for local processing, RENAL-SMART for uploads to the national sovereign Datacenter.
RETICE-RENAL Infrastructure →KAT-AI is available from the very first RETICE deployment. Included in the RETICEO BOO offer — zero initial investment, zero Internet dependency, total protection of educational data.