MOZN

From a fraud product to a platform its own engineers install

MOZN was long known for a single product. It now presents itself through the “MOZN AI Platform” and six solutions: predictive and decision intelligence, enterprise knowledge and automation, fraud and investigation, finance and budgeting, procurement and supply chain, and regulatory and governance intelligence. The About page still names OSOS, its earlier enterprise knowledge product, which appears to have been folded into the new knowledge and automation solution.

The delivery model matters more than the solution list. MOZN does not hand over software and leave. It sends a pod of “forward-deployed engineers” (AI engineers, data scientists, governance and infrastructure specialists) to work inside your organisation, build on your data, and then train your staff to run it. That is slower and more expensive than buying a SaaS tool, but it suits institutions without an in-house AI team.

FOCAL is the clearest part, and the easiest to evaluate

FOCAL by MOZN is the most mature product and has its own site, getfocal.ai. It covers three areas: KYC and customer due diligence with sanctions screening; anti-money-laundering through real-time transaction screening and monitoring; and enterprise fraud management using behavioural monitoring, device intelligence, account-takeover detection and promotion-abuse controls. In September 2026 MOZN announced that FOCAL’s screening solution won a 2026 Datos Impact Award for sanctions and watchlist screening.

FOCAL’s site offers a “free demo for 30 days”, which none of MOZN’s other solutions do. That makes it the sensible way to test the company before committing to anything larger.

Arabic and sovereignty: claims to test before you sign

MOZN calls its platform “sovereign by design” and says it runs inside your own environment or the cloud you choose, with you keeping control of models, agents and workflows. In September 2026 it announced partnerships with NourNet for cloud and data-centre infrastructure in Saudi Arabia, and with MDS and Ejada for integration and delivery.

On language, it promises native Arabic understanding, including scanned, dialectal and decades-old documents. That is exactly what to test rather than take on trust. During a proof of concept, give it twenty real documents from your archive: poorly scanned letters, minutes full of proper names, text that mixes Arabic and English. Ask questions whose answers you already know, and check whether the system cites the right page.

Pricing: nothing published

Neither mozn.ai nor getfocal.ai lists prices or plans; every route ends in a demo request or a sales conversation (checked October 2026). Expect cost to depend on project scope, the size of the engineering pod and the hosting model.

Who it suits

MOZN fits banks and financial institutions regulated by the Saudi Central Bank and its Gulf counterparts, and government bodies that need AI on Arabic data without it leaving their environment. Small businesses and individuals will find nothing here. If your priority is understanding what customers say rather than compliance and decisions, look at Lucidya. If you would rather build on a Saudi Arabic model yourself, read about ALLaM.

Official site: mozn.ai.