Customer profile
Every customer has a risk profile with a live risk score, a status and the full history of events and devices. Identity and document verification results join the same record.
Naiza by Limaz
Review fraud and AML decisions in one workflow.
Naiza connects identity, transaction, device, and screening signals to configurable decisions and investigator review.
One decision layer, from signal to action.
A decision begins with the signals attached to a customer or transaction and ends with a recorded action that an investigator can examine.
The problem Naiza solves
A payment alert may arrive without the onboarding evidence that explains the customer. A screening match may sit in a different queue from the transaction that triggered it. Investigators then rebuild the same timeline by hand.
Naiza keeps the relevant signals, rule result, reviewer action, and case evidence on one decision path. The institution retains control of its policies, thresholds, and escalation process.
Core product capabilities
- Fraud decisioning
- AML screening
- Transaction monitoring
- Identity and document verification
- Device and behavioral intelligence
- Rules, decisions, review, and case operations
- Integration through APIs and SDKs
Typical users and industries
Naiza is designed for banks, finance companies, payment service providers, fintechs, and digital lenders operating fraud and AML controls in Saudi Arabia and the Gulf.
Fraud and AML teams use the decisions and review queues; risk and compliance leaders define policy and oversight; technology teams own the event and system integrations.
Integration model
APIs and SDKs pass onboarding or payment events to Naiza and return the resulting action to the originating workflow. Review outcomes can then be recorded against the same decision context.
The product documentation on naiza.ai contains the current integration detail; Limaz delivers the surrounding operating model and implementation work separately.
Saudi and GCC operating context
Naiza is built by Limaz for institutions that must operate under local regulatory expectations, bilingual operations, and high accountability for decisions that affect customers and the institution.
Where counter-fraud requirements, AML programs, or identity controls need broader system work, Limaz can implement around the product rather than leaving a platform disconnected from the operating model.
How a decision is made
Every event follows the same path, from the moment your product sends it to the moment an analyst records the outcome.
Your product sends an event
A signup, login, payment, withdrawal or profile change reaches Naiza through the REST API. It carries the customer ID, device, IP address, session and any product fields your rules need. Use a synchronous call when the flow must wait for the answer, and an asynchronous call when it can continue.
Naiza adds context
Naiza enriches the event with IP geolocation (country, city and internet provider), VPN and proxy detection, email validation, the device's history and the customer's past events. When the Web SDK is installed, behavioural signals from the same browser session join the record.
Rules run in priority order
Your rules run in the order you set, and the first match wins. Score rules add points to a cumulative risk score. When the score reaches 100, the event goes to review.
Naiza returns an action
The response carries ALLOW, REVIEW or BLOCK, a risk score from 0 to 100 and the reasoning behind the result, including the rules evaluated. Your system acts on it at once.
Analysts confirm the outcome
Events sent to review appear in the dashboard. Analysts confirm the result and write it back as a label: fraud, legitimate, chargeback, refund, false positive or true positive, with notes and your own case ID.
Confirmed cases sharpen the rules
Labelled outcomes show which rules catch fraud and which stop good customers, so the team tunes rules and thresholds with evidence.
Signals Naiza brings together
Naiza is the Arab world's first unified fraud and AML platform. Fraud signals and anti-money laundering (AML) screening sit on the same customer record, so every decision sees the full picture.
Device intelligence
Device fingerprints persist across sessions. Naiza scores each device, flags any device linked to three or more customer accounts, and exposes fraud rings through repeated identifiers and device behaviour.
Behavioural signals
The Web SDK reads how a session behaves: keystroke timing, mouse movement and entropy, and clipboard pastes. These signals separate a real customer from a script or an account takeover.
Network and email
Every event is checked for IP location by country, city and internet provider, for VPN and proxy use, and for a valid email address.
Transactions and velocity
Payment, transfer and withdrawal events carry amounts, session and velocity context, and the product fields your rules evaluate.
Sanctions and watchlists
Customers and counterparties are screened against 1,100+ global sources and millions of profiles, covering sanctions, politically exposed persons (PEPs), their relatives and close associates, watchlists and adverse media.
Your own lists
Block and allow lists by email, device fingerprint or IP address. Each entry carries a reason and an optional expiry date, and blocked entries stop in real time.
AML screening across the customer lifecycle
- At onboarding
- Screen each new customer at registration using name, date of birth, nationality, entity type and known aliases, before the account opens.
- On every payment
- Screen the counterparties in payments and transfers, and send AML events for transaction and counterparty updates so risk is assessed continuously.
- Across the portfolio
- Re-screen the existing customer base continuously, so a new listing surfaces on the customer it affects.
- Arabic and Latin names
- Naiza normalises Arabic and Latin names and handles spelling variations, aliases and differences between scripts. One customer's name in several spellings resolves to one match decision.
- Clear outcomes
- Each screening returns AUTO_CLEAR, REVIEW or CONFIRMED with scored matches. A sensitivity profile sets how strictly names match, and REVIEW and CONFIRMED results hand off to your analysts.
Built for Saudi and GCC requirements
Naiza stores and processes fraud events, customer records and AML screening data inside the Kingdom of Saudi Arabia. API calls stay in the Kingdom, data in transit uses TLS, data at rest is encrypted, and access follows role-based controls.
The Saudi Central Bank (SAMA) Counter-Fraud Framework, issued in October 2022, organises counter-fraud work into four domains: Governance, Prevent, Detect and Respond. It expects institutions to run counter-fraud systems and verify they work as intended, document the rationale for scenarios and thresholds, add new rules quickly when new fraud types appear, and investigate and record the outcome of every system alert. Naiza's rule engine, decision reasoning and labelled outcomes give fraud teams the tooling for each of these expectations.
The Saudi Anti-Money Laundering Law requires financial institutions to apply customer due diligence, use appropriate systems to identify PEPs, monitor transactions on an ongoing basis, keep records for at least ten years, and report suspicious transactions promptly and directly to the General Directorate of Financial Intelligence (SAFIU). The Law of Combating Terrorism Crimes and its Financing sets matching duties for counter-terrorist financing (CTF). Naiza's screening, monitoring and audit trail give compliance teams the systems and records these duties call for.
Limaz built Naiza in Riyadh for regulated digital products in Saudi Arabia, with Arabic name matching at the core of its screening.
For fraud teams, AML teams and investigators
Fraud teams
Write and change rules without a deployment, test them before activation, and watch a new rule in observation before it enforces. Every event shows the reasoning behind its decision.
AML and compliance teams
Screen at onboarding, on payments and across the portfolio. Scored matches, clear REVIEW and CONFIRMED hand-offs and a full audit trail per screening support every escalation.
Investigators
Open a customer and see every event, device, decision and screening result in one timeline. Record the outcome with notes and your case ID, on the same decision.
Risk and compliance leaders
Set policy and thresholds, follow alert volumes and outcomes in the dashboard's analytics, and hold the evidence that examiners and auditors ask for.
Engineering teams
A REST JSON API, examples in cURL, Node.js and Python, separate test and live keys, and signed webhooks. Team access uses Admin, Analyst and Viewer roles.
Integration
- REST API
- JSON over HTTPS with API key authentication. Test keys and live keys keep development and production apart.
- Synchronous decisions
- Call the evaluate endpoint when the product flow needs an answer before it continues, such as a login or a payment approval.
- Asynchronous monitoring
- Send events in the background when the flow can continue, and receive results through webhooks.
- Web SDK
- Collects device and behavioural signals in the browser. A server-issued token links those signals to your own session and to the events that follow.
- Signed webhooks
- Callbacks for evaluated events, submitted feedback and blocklist changes, signed with HMAC-SHA256 and retried automatically with exponential backoff.
- Feedback API
- Write investigation outcomes back to a decision or an event, with a label, notes and your external case ID.
Case evidence and audit trail
- Decision record: the action, the risk score from 0 to 100, the rules evaluated and the reasoning, for every event.
- Event trail: every event kept for operations, feedback and audit.
- Customer history: every event and device linked to the customer's risk profile.
- Screening record: scored matches and the result for each screening, with a full audit trail.
- List changes: each block or allow entry carries a reason, creation and update times and an optional expiry, with an audit trail for every change.
- Investigation outcome: the confirmed label, analyst notes and your external case ID, attached to the original decision.
- Access control: role-based access for Admin, Analyst and Viewer users, and authenticated API calls.
Further reading
Saudi financial fraud law and SAMA's Counter-Fraud Framework: what institutions must do
The law punishes fraud after it happens. SAMA's Counter-Fraud Framework requires institutions to prevent and detect it first. Here is how the two fit together, and what a working programme looks like.
Saudi AML law for fintechs: building the programme SAMA expects
A fintech that moves, holds or lends money in the Kingdom is a financial institution under the AML Law. This is how to turn the law's obligations into a programme that works at fintech speed.
Questions
Naiza is Limaz's fraud and AML decisioning product. It brings screening, transaction monitoring, identity signals, rules, and review workflows into one layer for financial institutions in Saudi Arabia and the Gulf.
Banks, finance companies, payment service providers, fintechs, and digital lenders that need accountable financial-crime operations under local regulatory expectations.
Product capabilities, APIs, and integration guides live on naiza.ai. This Limaz page explains the parent product relationship and the surrounding delivery work.
Yes. Naiza is the Arab world's first unified fraud and AML platform. It scores logins, signups, payments and withdrawals for fraud, and screens customers and counterparties against sanctions, PEP and watchlist data, on the same customer record. Fraud analysts and AML analysts work from one timeline of events, devices, decisions and screening results, so a mule account or a sanctioned counterparty is visible to both teams at once.
Naiza stores and processes fraud events, customer records and AML screening data in the Kingdom of Saudi Arabia, on Saudi-hosted infrastructure. Fraud and AML API calls stay inside the Kingdom. Data in transit uses TLS, data at rest is encrypted, API calls are authenticated with API keys, and user access follows role-based controls.
Naiza screens customers and counterparties against 1,100+ global sanctions, PEP and watchlist sources in a single call, covering millions of individual and entity profiles. Screening covers sanctions, politically exposed persons (PEPs) and their relatives and close associates, watchlists, adverse media, entities and counterparties. Each screening returns AUTO_CLEAR, REVIEW or CONFIRMED with scored matches and a full audit trail.
Naiza normalises Arabic and Latin names and handles spelling variations, aliases and differences between scripts. A name written in Arabic on an ID and in Latin letters on a payment resolves to the same match decision. A sensitivity profile sets how strict matching is, and possible matches go to REVIEW with a score so analysts see why a name was flagged.
SAMA's Counter-Fraud Framework expects working counter-fraud systems, documented scenarios and thresholds, timely rules for new fraud types, and a recorded outcome for every alert. Naiza gives fraud teams an expression-based rule engine that changes without a deployment, rule testing before activation, decision reasoning on every event, and labelled outcomes with case IDs for every review.
A developer sends a first event in minutes with a test API key and the examples in cURL, Node.js or Python on naiza.ai. Production rollout follows your product flows: synchronous calls for decisions a flow must wait for, asynchronous events for background monitoring, the Web SDK for device and behavioural signals, and signed webhooks for results. Limaz delivers the wider integration work when you need it.
Yes. Naiza rules use expressions with full logical operators, run in a priority order you set, and switch on or off without a deployment. Teams test and validate a rule before activation and run it in observation first. Each rule has a human-readable name that appears in review and feedback, so analysts know which rule raised each case.
Discuss your requirements with Limaz
Tell us which system, control, or requirement you are working on. We’ll come back with a clear next step.