Webinar: AI-Driven Fraud Detection
1
Your Hosts
2
Lead Architect
Codit Luxembourg
Maxime Dehaut
Head Cloud Apps, Data & AI
Telindus Luxembourg
Frank Roessig
Agenda
| Context on AI-Driven Fraud Detection
| Use-Cases
| Outlook
3
Online Transactions
October 2020 Webinar: AI-Driven Fraud Detection 4
Fraud is a significant cost factor
October 2020 Webinar: AI-Driven Fraud Detection 5
Fraudsters using Increasingly Complex Strategies
October 2020 Webinar: AI-Driven Fraud Detection 6
Fraud touches many Sectors
October 2020 Webinar: AI-Driven Fraud Detection 7
AI/ML Adoption
October 2020 Webinar: AI-Driven Fraud Detection 8
Operationalised
Models 20% Adopters80%
Model
Management
Business driven
initiatives
Company Culture
25%
40%
Skills & Expertise
55%
50%
Source: Gartner, McKinsey, O’Reilly – Reports from 2019 & 2020
Fraud Management
October 2020 Webinar: AI-Driven Fraud Detection 9
Prevention
Treatment
Detection
Recovery
Prevention & Detection
October 2020 Webinar: AI-Driven Fraud Detection 10
Data
Operations
Practices
Functionalities
October 2020 Webinar: AI-Driven Fraud Detection 11
PracticesData
Operations
ENRICHMENT
INGESTION
VISUALISATION
EXPLORATION
EXPLAINABILITY
GOVERNANCE
SECURITY
MONITORING
EXECUTION
AUTHORING
CASE MANAGER
PIPELINE
ANALYTIC ENGINE
RULE ENGINE
Augmented Rule Engine
October 2020 Webinar: AI-Driven Fraud Detection 12
As a ServiceUncertainty
Accountability
& Ownership
Feedback loop
ML-driven Fraud Management System
October 2020 Webinar: AI-Driven Fraud Detection 13
MATURITY
SOPHISTICATION
AI/ML
Platform
CTI
Explainability
Fairness & Diversity
Governance
Robustness &
Security
Rule-based &
Analytic
Models
Cognitive &
Probabilistic
Analytics
Reference case
14
AI Design
Experience
Transform your AI vision into reality with this
jumpstart program
| Finetune your AI business case during an
interactive workshop
| Sharpen your AI vision with real-world use
cases
| Demonstration of how an AI solution is
operationalized with MLOps
| First Proof of Concept with your actual
data
| Insights on the Azure PaaS architecture
Thanks for Attending
15

AI-Driven Fraud Detection

  • 1.
  • 2.
    Your Hosts 2 Lead Architect CoditLuxembourg Maxime Dehaut Head Cloud Apps, Data & AI Telindus Luxembourg Frank Roessig
  • 3.
    Agenda | Context onAI-Driven Fraud Detection | Use-Cases | Outlook 3
  • 4.
    Online Transactions October 2020Webinar: AI-Driven Fraud Detection 4
  • 5.
    Fraud is asignificant cost factor October 2020 Webinar: AI-Driven Fraud Detection 5
  • 6.
    Fraudsters using IncreasinglyComplex Strategies October 2020 Webinar: AI-Driven Fraud Detection 6
  • 7.
    Fraud touches manySectors October 2020 Webinar: AI-Driven Fraud Detection 7
  • 8.
    AI/ML Adoption October 2020Webinar: AI-Driven Fraud Detection 8 Operationalised Models 20% Adopters80% Model Management Business driven initiatives Company Culture 25% 40% Skills & Expertise 55% 50% Source: Gartner, McKinsey, O’Reilly – Reports from 2019 & 2020
  • 9.
    Fraud Management October 2020Webinar: AI-Driven Fraud Detection 9 Prevention Treatment Detection Recovery
  • 10.
    Prevention & Detection October2020 Webinar: AI-Driven Fraud Detection 10 Data Operations Practices
  • 11.
    Functionalities October 2020 Webinar:AI-Driven Fraud Detection 11 PracticesData Operations ENRICHMENT INGESTION VISUALISATION EXPLORATION EXPLAINABILITY GOVERNANCE SECURITY MONITORING EXECUTION AUTHORING CASE MANAGER PIPELINE ANALYTIC ENGINE RULE ENGINE
  • 12.
    Augmented Rule Engine October2020 Webinar: AI-Driven Fraud Detection 12 As a ServiceUncertainty Accountability & Ownership Feedback loop
  • 13.
    ML-driven Fraud ManagementSystem October 2020 Webinar: AI-Driven Fraud Detection 13 MATURITY SOPHISTICATION AI/ML Platform CTI Explainability Fairness & Diversity Governance Robustness & Security Rule-based & Analytic Models Cognitive & Probabilistic Analytics
  • 14.
    Reference case 14 AI Design Experience Transformyour AI vision into reality with this jumpstart program | Finetune your AI business case during an interactive workshop | Sharpen your AI vision with real-world use cases | Demonstration of how an AI solution is operationalized with MLOps | First Proof of Concept with your actual data | Insights on the Azure PaaS architecture
  • 15.