Covariance logo

Covariance

Data-driven AI for insurance fraud detection and claims management

Claims Technology Startup Bootstrapped
Visit Website

Overview

Covariance is a fraud detection and claims analytics platform for insurance carriers and healthcare organizations, operating primarily in the Greek and broader European markets. Founded in 2019 and headquartered in Attica, Greece, the company develops AI-driven software that automates claims review workflows, identifies potentially fraudulent submissions before reimbursement, and supports cost reduction for P&C and health insurers.

The company offers two core insurance products -- CO Assess and CO React -- delivered as SaaS platforms that integrate with existing insurer systems. CO Assess applies machine learning to structured claims data for pre-payment fraud detection, while CO React combines multiple AI components (ML models, image similarity, network analysis, financial benchmarking, and mobile metadata) to address car insurance fraud specifically. Both platforms feature no-code dashboards with embedded visualization. The company also provides autonomous AI agent solutions for workflow automation and general advanced analytics services for the insurance and healthcare sectors.

Covariance reports 11-50 employees with over 75% holding MSc or PhD credentials. Notable engagements include a project supporting the Cypriot Health Insurance Organisation via KPMG Cyprus in 2025 and participation in the EU Horizon Europe CyberSecDome programme in 2026. The company holds ISO 9001:2015 and ISO 27001:2013 certifications and is a member of the NVIDIA Accelerator Hub (first in Greece).

Products & Services

CO Assess

An AI-driven fraud detection platform for insurance carriers that integrates with existing claims management systems. Uses machine learning on structured claim data to flag potentially fraudulent claims before payment, reducing loss ratios and claims costs.

Key Features

  • Pre-payment fraud scoring via machine learning on structured claims data
  • Integration with existing insurer systems and databases
  • Subscription-based pricing scaled to claims volume
  • Deployable and operational within weeks
  • Dynamic dashboards and real-time reporting

Target Users: P&C and health insurance carriers

CO React

A multi-technology fraud detection platform targeting car insurance claims, combining several AI components to assess and flag fraudulent submissions before reimbursement.

Key Features

  • ML data models for structured claim fraud scoring
  • Deep learning image similarity to detect repeat accident photos and manipulated images
  • Network analysis to map fraud rings involving individuals and service providers
  • Financial data analysis benchmarking claims against historical averages
  • Optional mobile metadata analysis for driver behavior via dedicated app
  • No-code operation with dynamic dashboards and embedded visualization

Target Users: P&C auto insurance carriers

Advanced Analytics & Machine Learning

General analytics and machine learning services for insurance and healthcare organizations handling large, complex datasets, including predictive analytics, claims analytics, risk analysis reporting, and pricing/underwriting support.

Key Features

  • Predictive analytics and risk scoring
  • Claims trend analysis and reporting
  • Underwriting and pricing support
  • Custom ML model development

Target Users: Insurance carriers and healthcare organizations

Autonomous AI Agents

Agentic AI solutions for workflow automation and virtual operations within insurance and healthcare environments.

Key Features

  • Automated workflow execution
  • Virtual operations for claims and administrative processes

Target Users: Insurance carriers, healthcare organizations