Meelo

Cross analysis of 400 signals per individual, spanning document authenticity, biometrics and behavior in the digital journey, returns a decision support score in two to five seconds from proprietary evolving AI. Manual file review across disconnected tools would be all that remained.

Critical alerts help the analyst prioritize rather than dictate the procedure, and analysts keep decision authority, with adapted controls for entrepreneurs and young companies. Thresholds, and what proceeds automatically where the product replaces an existing process, are not described.

Each score comes with an explanation of why it was given and which signals raised the alert. Accuracy, false positive and detection rates are unpublished, and no supervisory mapping for the models is given.

Arkéa Financements Services reports fraud risk divided by twelve, with no period or method. More than 60 clients include Crédit Agricole, Ayvens and AXA, and unnamed clients report 1.8 million euros saved over two years.

Banks, lenders and leasing groups competing for the same customers share the platform, and fraud detection typically improves by observing outcomes across all of them. Whether one client's fraud pattern informs scoring at another, and whether results are retained or reused, is not stated.

Full GDPR compliance is claimed, and an external audit is said to have rated privacy compliance excellent. The auditor, retention for verification and bank data and the subprocessors are not named.

Encrypted storage and a business continuity plan are described, beside B Corp and EcoVadis Gold, which are sustainability credentials. An information security certification or trust portal appears nowhere on the pages.

The GDPR, PSD2 for bank data access, the new AML regulation and sixth AML directive, and eIDAS 2.0 are named design targets, with PSD3 and FiDA ahead. France's banking supervisor never appears.

Algorithms are audited against an IEEE standard for fair, transparent and explainable AI, and a team supervises model performance and machine bias. Audit results, the auditing body and acceptance rates by population are not published.

Score explanations let an analyst reconstruct, justify or overturn a decision, but no guarantee or accuracy commitment stands behind them. Applicants rejected over a poorly scanned document or an atypical profile are not told which signal caused it and are offered no route to contest.

Infogreffe and Pappers supply official company data, and the models are described as proprietary. Neither the watchlist source nor the biometric and document engines are identified, which matters for a sovereignty pitch.

A portal, native API or batch processing enriches existing journeys, with Powens as open banking provider and deployment in about four weeks. No core banking, loan origination or onboarding system is named, and API documentation is not linked.

Data stays on Microsoft Azure in France with no transfer outside European territory. Its platform is offered only as SaaS, with no private or on premises option.

Consolidating checks otherwise bought from several vendors is the stated cost argument, yet nothing is priced. Per check, per verified customer or subscription charging, and the relative price of each module, are never given.

Individual and business verification, AML screening and solvency assessment sit in one platform, serving banks, leasing and mobility groups and insurers. Coverage runs across eight countries, while dedicated material by institution type is thinner than the product pages.