The Future of Data Analysis Combines
Entity Resolution with Graph Visualization

Find the Best Customers & Worst Criminals Hidden in Your Data

Some relationships in data remain invisible until you have the proper tools to see them. At the 2025 Gartner Data & Analytics Summit, Matthieu Besozzi of Linkurious and Paco Nathan of Senzing demonstrated how combining graph analytics with powerful entity resolution technology reveals connections that would otherwise remain hidden in your data.

In their presentation, Paco and Matthieu explain how the Linkurious Decision Intelligence platform, now with entity resolution fully integrated with Senzing Inside™, helps organizations uncover valuable insights about their customers while identifying potential fraud – all through a business-friendly interface that doesn’t require data science expertise.

Fraudsters deliberately conceal connections through techniques like “channel separation,” a tradecraft where criminals gradually alter personal information across multiple accounts. As Paco Nathan explains, “By doing channel separation, Bob Jones is going to set up other accounts. His personal information will change just a little bit, so the data drifts.”

Eventually, a fraudster establishes accounts with entirely different identities, making traditional detection methods ineffective. This issue extends beyond criminals to everyday customers. Matthieu Besozzi illustrates this challenge using himself as an example:
“My name is Matthieu Besozzi. I have a first name that is French. A last name that is Italian. Here in America, everybody calls me ‘Matthew Besozzi.’ Some people use ‘Mathew Besozzi’ with just one ‘t.’ My favorite name is actually my name on my passport: Matthieu Stefan Bernard. These [disconnected] identities are actually the same person.”

For financial institutions, disconnected identities create a major obstacle when trying to form a complete understanding of entities and their relationships – impacting AML compliance (KYC, Customer Due Diligence, AML Investigations), fraud detection and forensics analysis.

Linkurious’ platform efficiently and smoothly unifies data from multiple sources into one single source of truth, and generates an entity resolved knowledge graph without hassles. It provides a cleaner, clearer, more reliable view of complex data and reveals hidden relationships between millions of entities. With a comprehensive and accurate view of connected data at scale, analysts and investigators can uncover critical contextual insights for fraud detection, risk management and customer relationship optimization.

During the presentation, Besozzi illustrates a common situation where a system might show two clients who appear unrelated – one regular customer and one flagged as suspicious – when they’re actually the same person using slightly different credentials. Without proper entity resolution, these critical connections remain unknown.

During the Gartner event presentation, Besozzi reveals why Linkurious chose Senzing as their entity resolution partner: “What I find interesting with Senzing is explainability” – which is because for each node that is resolved, Senzing explains explicitly why or why not, as well as how every decision was made.

In addition to explainability, Besozzi highlights two other critical factors: “I’ve seen a lot of entity resolution tools in which data is actually deleted or changed, and you cannot go back.” This never happens with Senzing. The technology’s non-destructive approach preserves data integrity while enhancing connectivity.

Finally, Besozzi emphasizes performance: “Speed and scalability are very important. Senzing can work on huge datasets – billions of entities can be resolved very fast.”

The live demonstration at the Gartner summit showcased how seamlessly Linkurious with Senzing Inside work together. “This is the first time we are really showing the public our end-to-end integration with Senzing,” Besozzi explains. The demonstration follows a straightforward workflow where Besozzi maps entities, activates entity resolution with a single click and shows how the system quickly identifies potentially duplicated entities throughout the graph.

When searching for a specific person in the Pandora Papers dataset, Besozzi finds three separate entity records that Senzing had flagged as the same individual. “Now that I’ve resolved the entity, I can merge all these people into one super-node that is representing the person,” he explains. The technology is even more impressive at scale. Besozzi demonstrates how a complex graph with numerous disconnected entities suddenly reveals its true structure after entity resolution is applied.

To learn more about how the Linkurious Decision Intelligence platform with Senzing Inside™ can reveal critical connections in your data, visit Linkurious.com and explore their graph visualization and investigation platform.

For those who want to dive into specific aspects of the presentation, here are key moments:

00:18 Origins of Senzing Entity Resolution Technology
Introduction to entity resolution and how it evolved from Jeff Jonas’s work in Las Vegas casinos developing “non-obvious relationship awareness.”

03:39 Entity Centric Learning Explained
How Senzing entity resolution’s approach differs from traditional record-to-record comparisons to reveal connections that would otherwise remain hidden.

05:39 Cultural Context in Entity Resolution
Demonstration of how Senzing handles complex name variations across cultures and languages with 80 different globalizations.

08:17 Introduction to Linkurious Enterprise
Overview of the graph visualization and investigation platform designed specifically for business users.

10:06 Real-World Identity Challenges
Matthieu Besozzi’s personal example of how his name appears in multiple forms across different systems.

13:12 Live Entity Resolution Demo
Step-by-step demonstration of the entity mapping process in Linkurious with Senzing integration.

14:32 Practical Application with Real Data
Using Pandora Papers data to demonstrate how multiple entity records for the same person can be identified and merged into a comprehensive “super node.”

16:24 Why Linkurious Chose Senzing
Explanation of the key factors: explainability, non-destructive processing and unmatched scalability.

17:33 Integration Benefits for Massive Datasets
How the combined solution delivers particular value for organizations dealing with billions of entities where manual review would be impossible.