Real-World Anomaly Detection AI Use Cases and How To Get Started

Detecting anomalies at scale is a critical challenge for data-driven organizations. Traditional rule-based systems often fail to adapt to evolving patterns, resulting in false positives, missed fraud, and costly inefficiencies. For practitioners building AI/ML solutions, these limitations slow innovation and reduce model accuracy.

Teradata sees a range of anomaly detection applications through our AI pilot program—from data quality monitoring and fraud prevention to manufacturing optimization. Real-time detection of these anomalies can be imperative to prevent downstream impact.

ClearScape Analytics® provides the flexibility and performance needed for modern AI use cases—enabling teams to streamline the entire process from data preparation and exploration to model training and deployment. This includes going beyond raw transactional features to capture the deeper semantic relationships in the data through open-source language models with Bring Your Own Model (BYOM) capabilities.

Watch the replay to see anomaly detection use cases from pilot to production and learn how to start experimenting yourself. In this session, you'll see how to:

• Transform transactional data into rich embeddings using ONNX models directly within ClearScape Analytics®, capturing patterns that traditional features miss
• Apply in-database K-means clustering to identify anomalous transactions at scale without moving data out of Teradata’s platform
• Use the elbow method and silhouette scores to systematically determine optimal cluster configurations
• Visualize and interpret results to reduce false positives and surface meaningful fraud patterns 

Presenter(s):

Matt Mazzarell

AI Lead, Financial Services, Americas

Matt Mazzarell is AI Lead for Financial Services in the Americas, where he helps customers leverage Teradata’s AI capabilities to solve complex, data intensive business problems. Matt has helped some of the largest banks improve customer experience and promote greater operational efficiency. He specializes in leveraging unstructured text data with language models that run in-database to promote optimal production performance for AI workflows. He utilizes experience and business acumen to make sure AI is applied in the most business relevant context.

Daniel Herrera

Principal Developer Advocate, Teradata

Daniel Herrera is Principal Developer Advocate at Teradata, specializing in AI/ML and advanced analytics. He helps developers and data practitioners unlock enterprise-scale innovation through ClearScape Analytics and VantageCloud. With deep expertise in data engineering, cloud architecture, and model deployment, Daniel bridges technical complexity with practical solutions. A frequent speaker and community contributor, he focuses on enabling scalable AI use cases across industries.