Yaqaza · Pitch Deck
Banking & Fintech · REAL-TIME FRAUD
Yaqazaيقظة
A stream-native fraud platform that decides in milliseconds and never stops watching.
neuriya.ai — Sovereign, Arabic-first enterprise AI
Problem
Fraud models score against duplicated data, and most decisions land after the money has already moved
- Device, behavioral, and network features are frequently computed separately for each fraud model, so different systems see slightly different signals for the same transaction.
- Many fraud checks still run after settlement or with enough lag that a confirmed bad transaction has already cleared before anyone can act on it.
- Rules built around single accounts miss coordinated fraud rings that spread activity across linked accounts, devices, and beneficiaries.
- Confirmed fraud cases are typically packaged into SAMA regulator reports by hand each cycle, a manual assembly job that competes with actual investigation time.
Solution
One shared feature store, one streaming decision, before the transaction settles
Product
Yaqaza in action
How it works
From first touch to outcome
Stream
Transaction, device, and behavioral events flow into the shared feature store in real time from every channel.
Score
The decision engine scores and actions each transaction in milliseconds, before it settles.
Investigate
Flagged cases land in a prioritized workspace with full evidence and linked-account context attached.
Report
Confirmed cases are batched and submitted to SAMA automatically, on schedule, every cycle.
Market
Sizing the opportunity
- TAM — $2.1B. GCC real-time fraud detection & case management market
- SAM — $230M. Saudi bank & fintech real-time fraud platform market
- SOM — $16M. 3-year obtainable share across licensed Saudi banks and fintechs
SAMA's Open Banking rollout and the shift to instant payment rails are pushing fraud detection from a batch, post-settlement process to real-time infrastructure that licensed institutions are expected to have in place, not a discretionary upgrade.
Business model
How Yaqaza makes money
Usage-based transaction scoring
Metered pricing per transaction scored through the streaming decision engine.
Case management platform subscription
Recurring fee for the investigator workspace, linked-account graph, and regulator-submission tooling.
Implementation & feature-store integration fee
One-time fee to connect transaction, device, and behavioral data sources into the shared feature store.
Revenue streams
Where the revenue actually comes from
Competition
Where we win
Our edge
- Sub-40-millisecond streaming decisions before settlement, not a batch process that runs after the transaction has already cleared.
- Fraud-ring graph detection catches coordinated, linked-account activity that single-account rules miss entirely.
- Automated SAMA report packaging removes the manual assembly work that competes with actual investigation time.
- Founding team combines Saudi enterprise delivery experience with hands-on applied-AI and streaming-systems engineering.
Traction
Early validation, built for pilot readiness
- Full real-time fraud platform designed and built end-to-end: fraud KPI dashboard, flagged transaction queue, and fraud case detail.
- Architecture pre-aligned to SAMA CSF, PDPL, ISO 27001, and NDMO Classification.
- Actively selecting design-partner banks and fintechs for a first pilot cohort.
Go-to-market
Land on one channel, prove the false-positive gain, expand to full transaction flow
- 1Land with 1-2 design-partner banks or fintechs, starting on a single high-volume payment channel scored against their live traffic.
- 2Prove reduced false positives and faster time-to-decision over a defined pilot period against the institution's existing fraud rules.
- 3Expand across the institution's full transaction flow, then through inter-bank fraud-network relationships to reach the broader licensed-institution market.
Team
Built by operators, not just builders
The ask
Raising a seed round to reach first design-partner banks and fintechs
We're raising to fund streaming decision-engine hardening, core-banking and payment-rail integrations, and the core data-science and engineering team through first-revenue pilot deployments.