Datavillage

Confidential AI collaboration for fraud detection and AML compliance in financial services.

Claims Technology Startup Seed

Overview

Datavillage is a fraud and AML intelligence platform for banks and financial institutions, including insurers. Founded in 2019 and headquartered in Brussels, Belgium, the company enables organizations to detect financial crime by combining sensitive data across institutional boundaries using confidential computing technology.

The platform provides encrypted collaboration environments -- referred to as confidential collaboration spaces -- where participants can jointly train and apply AI models on combined transaction and behavioral data without exposing underlying records to any party. Data remains encrypted during processing via secure enclaves (trusted execution environments), and no raw data is shared between participants. The approach addresses cross-institutional fraud that evades detection when organizations work from data silos alone. Key products include the core platform and Mule Monitor, an AI agent for detecting mule account activity.

Datavillage raised approximately EUR 1M in seed funding in 2021 from investors including SambrInvest, Stadiem, and Spreds, with additional EU Horizon 2020 grant support. The company is partnered with GraydonCreditsafe for collaborative fraud prevention, and is listed as a solution provider in the Temenos Exchange Partner network. A 2024 pilot with GraydonCreditsafe demonstrated the ability to detect an additional 10-20% of fraud cases beyond single-institution models.

Products & Services

Confidential Data Collaboration Platform

A platform enabling organizations to govern and operate confidential collaboration spaces where participants can train and run AI models on combined sensitive datasets without sharing raw data. Built on confidential computing (secure enclaves / encrypted data-in-use), it is cloud-agnostic and supports deployment on AWS, Azure, and GCP. Use cases include cross-institutional fraud detection, AML compliance, and open finance data initiatives.

Key Features

Data encrypted throughout processing -- no raw data exchange between participants
Full auditability of all collaboration activity
Regulatory compliance support for open finance and data sovereignty requirements
REST API integration with existing fraud and AML workflows

Target Users: Fraud prevention teams, compliance teams, and data partnerships teams at banks, credit information companies, and financial institutions

Mule Monitor

An AI agent for fraud and AML teams focused on detecting mule account activity. The product automates pattern detection in transaction data, reduces false positives, and provides investigators with faster response capabilities. It runs without data leakage to external LLMs and includes full auditability to support compliance requirements.

Key Features

AI-driven mule account detection
No data leakage to LLM providers
Real-time investigation support and pattern identification
Auditability for regulatory compliance

Target Users: Fraud and AML compliance teams at banks and financial institutions

At a Glance

Founded: 2019
Headquarters: Brussels, Belgium
Employees: 11-50
Funding: Seed

Category & Focus

Category: Claims Technology
Subcategories: Fraud Detection AML/KYC Confidential Computing Data Collaboration
Insurance Verticals: P&C Commercial Specialty/E&S
Target Customers: Carriers, Reinsurers

Customers

GraydonCreditsafe -- partnership for collaborative fraud prevention and risk management (announced August 2024)
RTBF (Belgian public broadcaster) -- early data collaboration customer
VRT (Flemish public broadcaster) -- early data collaboration customer