We design, build and transfer sovereign platforms for mobile coverage and quality-of-service compliance, device identity registers (CEIR), international voice gateway control and revenue assurance, and we build the AI that runs inside them. A British company whose founding partners have each spent more than twenty years delivering national programmes across Africa, the Middle East and Central Asia.
Analitico International brings under one roof partners who have each spent more than two decades delivering telecom regulatory platforms, international voice gateways and analytics across Africa, the Middle East and Central Asia. The references on this site were earned by that founding team. Our deepest roots are in telecommunications regulation; the same engineering discipline now serves banks, governments, hospitals and enterprises.
Coverage and QoS compliance, CEIR, international gateways, traffic and tariff control, revenue assurance.
Network analytics, fraud detection, QoS improvement tooling and regulatory reporting.
Sovereign data platforms, biometric enrolment, tax collection systems and digital transformation programmes.
AI strategies and roadmaps, agentic automation, risk and forecasting models for banks and investment funds in the GCC.
Diagnostic models, synthetic data generation and executive AI training programmes.
Both platforms are delivered on-premise, on open technology, with full source code and model transfer. No recurring licences, no vendor lock-in, no dependency on third-party public AI.
A national GIS platform that turns operator network data, drive tests and crowdsourced measurements into one trusted view of mobile coverage and quality of service. The regulator sees where licence obligations are met, where they are not, and which interventions move the most people.
A national register of every mobile device on the networks, keyed by IMEI. The regulator controls which devices may connect, blocks stolen, counterfeit and cloned handsets, enforces type approval and customs duty, and gives citizens a simple way to check a phone before they buy it.
| IMEI | Brand / model | Specs | Operator | First seen | Status |
|---|
Select any block to see what it does and why it is there.
Northbound counters, alarms, cell configuration and site registers from each operator's network management systems, delivered over secured links and validated on arrival.
INTEGRATION · SFTP / API · PER OPERATORPress play. Captions follow the language of the page.
Each opens on top of this page and runs entirely in your browser. Nothing you type is sent anywhere.
We are an AI company as much as a telecom company. Our AI practice covers the full cycle: strategy and roadmaps, executive and technical training, solution design, and the engineering of agents, orchestrators and domain "brains" that run inside our platforms and our clients' operations.
Plans the work, assigns tasks to specialist agents, checks their results and escalates to a human when confidence is low or policy requires approval. Every decision is logged.
ROLE · PLANNING · DELEGATION · AUDIT CLICK A NODE · SIGNALS SHOW AGENTS EXCHANGING TASKS AND RESULTSNational, sectoral and corporate AI strategies with governance, use-case portfolios, data foundations, operating models and investment roadmaps. Delivered for banks, funds, ministries and hospitals in the GCC.
More than 200 agents built and deployed: task agents, multi-agent orchestrators that plan and delegate, and agentic modules that plug into existing systems with guardrails, audit trails and human approval where it matters.
Domain knowledge cores built on open-weight language models with retrieval over the client's own regulations, contracts and data. Arabic-capable, deployable on-premise and air-gapped, with no dependency on public AI services.
Detection, classification, OCR and video analytics for inspection, security, document processing and field operations, trained on the client's imagery and tuned for edge or data-centre deployment.
Privacy-preserving synthetic datasets for training and testing, plus forecasting models for demand, risk, traffic and operations, with scenario simulation.
Board, executive and practitioner programmes in AI, strategy and transformation; more than 2,000 leaders trained in Qatar, Saudi Arabia and the wider region, delivered by faculty who also teach at leading business schools.
A vision language model is an AI model that understands images and video together with text. Classic computer vision answers one fixed question ("is there a crack in this mast?"). A VLM can look at a photo, a video frame, a scan or a document and answer open questions in natural language, read the text it contains, explain what it sees and hand the result to an agent.
In medicine, for example, a VLM reads a chest X-ray, highlights the region that concerns it, writes a plain-language finding with a confidence level and routes it to the radiologist for the decision. The same mechanism reads a tower photo, a customs form or a type-approval certificate for a regulator.
Each module stands alone or plugs into the same data fabric, so a regulator can start with one mandate and extend without re-platforming.
Monitoring and control of international voice traffic, grey-route and SIM-box detection, and verified settlement for the regulator. Deployed in seven African markets.
Independent measurement of operator traffic and tariff compliance against licence conditions and consumer-protection rules.
Reconciliation of declared versus measured revenue, automated levy computation and audit-ready reporting for ministries of finance and regulators.
From strategy to production: agents, multi-agent orchestrators, agentic modules and sovereign domain brains. More than 200 agents live across the Middle East and GCC.
Image, video and document understanding for inspection, security and compliance, with vision language models that read, explain and act.
End-to-end enrolment solutions and certified biometric devices developed with international market leaders.
AI, strategy, technology and leadership programmes for regulators, ministries, banks and funds; more than 2,000 executives trained.
Dev, test and production environments on the client's infrastructure, air-gap capable, with benchmarked compute, disaster recovery and multi-year support.
Full source code, models, documentation and training handed over, so the client operates, extends and audits the platform independently.
National telecom programmes delivered over fifteen years for regulators and, in most of the African markets, directly with mobile operators; AI strategies, agents, models and training across the Middle East.
Gateway monitoring, anti-fraud and settlement assurance delivered with mobile operators, and with regulators where mandated, from West Africa to the Great Lakes.
Consultancy, architecture and technical advisory for the Tajikistan regulator's QoS platform; national QoS and coverage monitoring programme in Oman.
Comprehensive technical audit for Chad's regulator, ARCEP, delivering actionable findings to the Director General.
AI strategies, agents, prediction and vision models, and executive training in Qatar, Saudi Arabia and Lebanon; 200+ agents built and 2,000+ executives trained.
Signed reference letters and verification contacts are available to evaluation committees on request.
"The comprehensive technical audit conducted proved instrumental in uncovering critical insights and actionable data. Thank you for your exceptional professionalism and a job very well done."
"We were able to improve our Quality of Service and provide better services to our customers thanks to your modules and your analytics tools."
"They provided continuously high quality services, and provided very efficient tools for fraud detection."
Analitico International is a member of the Microsoft AI Cloud Partner Program and of Google Cloud Partner Advantage.
Our platforms run fully on-premise by default. These programmes give our clients access to Microsoft and Google cloud and AI technologies when they choose a cloud or hybrid deployment.
Client names are withheld; full details and verification contacts are shared with evaluation committees under confidentiality.
A regulator needed to verify operators' coverage claims and enforce licence obligations with evidence that would hold.
International voice revenue was leaking through grey routes and SIM boxes, and settlements could not be verified.
A bank wanted AI beyond pilots: a governed strategy and agents that work inside its real processes.
A governance white paper for telecom regulators, policymakers and ministries: a reference architecture, a governance model and a practical roadmap for modernising quality-of-service oversight with AI, verifiable evidence and cross-sector coordination.
Reading guide: ministers and board members read about 10 pages (executive summary, costs, recommendations); technical, legal and operator teams each have their own path through the paper.
Charara, R. W. and Chatila, K. M. (2026). STORES: Sovereign Telecom Orchestration, Regulation and Enforcement System. Governance white paper, version 1.0, October 2026.
Stolen, counterfeit and smuggled handsets cost a country tax revenue, consumer safety and network quality. A CEIR closes all three gaps at once.
Every mobile device carries a unique IMEI. A Central Equipment Identity Register keeps the national list of IMEIs allowed on the networks, synchronised with each operator's equipment register. When a phone is reported stolen it is blocked on every network within minutes, which removes the resale value and, in every country that has deployed a CEIR, has reduced handset theft. Counterfeit and cloned devices, which interfere with radio performance and carry no safety certification, are identified by their invalid or duplicated IMEIs and placed in a grace period before blocking, giving owners time to regularise.
The fiscal side is often the decisive argument. Linking the register to customs means every imported device is declared and duty paid before it is activated, and the import channel moves from grey to formal. Regulators typically see the register financed by duty recovered in the first year, with the citizen-facing check service (SMS, web, app) building public support because buyers can verify a phone before paying for it.
All three, when they are reconciled on one platform. Each source alone can be challenged; together they produce evidence that holds.
Operators report coverage from their planning tools, which are optimistic by design. Drive tests measure what a car on a road experienced on one day. Crowdsourced measurements from citizens' phones are continuous and everywhere people live, but uneven in density and device quality. A regulator that relies on one source will lose the argument with the operator who relies on another.
The answer is a national platform that ingests all three, applies one agreed methodology (sampling rules, statistical confidence, indoor versus outdoor, peak versus off-peak) and computes licence compliance on a map the operator can inspect. Where sources disagree, the platform flags the area for targeted field verification instead of a nationwide campaign, which cuts measurement cost. The compliance finding then carries its evidence chain with it: which measurements, which method, which version of the model. That is what survives an operator's challenge, and it is also what lets the regulator direct universal-service funds to the places where people actually lack service.
Yes. Open-weight models, on-premise deployment and full transfer of code and models are now a realistic requirement, not a wish.
Regulators handle licence contracts, operator data and citizens' measurements. Sending any of it to a public AI service is usually prohibited and always unwise. The alternative used to mean no AI at all. It no longer does. Open-weight language and vision models now run on a few GPUs inside the regulator's own building, air-gapped if required, and are fine-tuned on the regulator's own documents so that a compliance officer can ask, in Arabic or English, "which sites are outside their licence obligation in this governorate and why?" and receive an answer with the evidence attached.
Three contract terms make this real: the models and source code are transferred to the regulator, the compute is benchmarked rather than tied to one brand, and the delivery includes the training of the regulator's own engineers to retrain and extend the models. With those in place the regulator owns the capability, not a subscription, and its AI roadmap is its own.
On-premise, air-gap capable, no reliance on public AI services. The client's data never leaves the client's infrastructure.
Open-source foundations, source code and models handed over, zero recurring licence fees.
A resident project director on site, workshops and documentation in the client's language, and knowledge transfer built into the plan.
Methodology aligned with ITU and GSMA practice, auditable computations and tamper-evident findings that stand up to operator challenge.
Methodology, KPIs and regulatory framework agreed with the client; site survey and hardware order placed early.
On-premise environments stood up; core platform, GIS and data fabric live for client users.
Operator feeds, field and crowdsourced data connected; AI models trained on the client's own data.
Field verification, security hardening, user acceptance and role-based training for the client's teams.
National launch, source code and model transfer, documentation and multi-year support.
For tenders, references, partnerships or training programmes.
Questions about this policy: info@analitico.tech.
