Anomaly detection

Minimize risks in your customer-facing, manufacturing, and IT management processes using machine learning.

Our clients

Retail
Hi-tech
Manufacturing
Finance
Healthcare
How our anomaly detection solutions work
Autonomous anomaly detection using unsupervised algorithms

We extensively use unsupervised machine learning algorithms to detect outliers and anomalies even in settings where “normal” metric patterns and thresholds are unknown. This eliminates the need for manual threshold management, enables the detection of complex multi-metric patterns, and allows scale-out to dozens of use cases and thousands of metrics.

Ready for extreme volumes of data
Our solutions are designed to monitor a very large number of metrics in real time, detect unusual patterns that involve multiple metrics at the same time, and automate root cause analysis at scale.
Simplified root cause analysis
We integrate our solutions with incident management and alerting systems to automatically send notifications and create tickets for operations teams. Notifications include most relevant segments of metrics that are potentially related to the incident. This helps to quickly react to incidents, reduce the amount of time needed for investigation, and prevent failure propagation or security-related losses.
Introduction to Algorithmic Marketing
Book
Introduction to Algorithmic Marketing
Introduction to Algorithmic Marketing is a comprehensive guide to advanced marketing automation for marketing strategists, data scientists, product managers, and software engineers.
We develop anomaly detection solutions for several industries including retail, ecommerce, manufacturing, technology, and video games
Video games
Virtual currency fraud and security and stability issues are common concern for video game developers and publishers. We help our clients in these industries to continuously monitor metrics collected from applications and to isolate and analyze issues.
Learn more about solutions for your industry

Read More

Unsupervised real time anomaly detection
In this article, we present a solution for real-time automated outlier detection in metrics collected from large-scale distributed systems (e.g., clusters of ecommerce applications) or data centers. We focus on handling a large number of metrics and detecting anomaly patterns in combinations of metrics rather than individual time series.
Read more
Add anomaly detection to your data with Grid Dynamics accelerator
In this article, we describe our real-time cloud-based anomaly detection reference solution. We cover its design and applicability to the most common use cases: monitoring, root cause analysis, data quality, and intelligent alerting. The solution is AI-driven and implements a flexible approach based on normal state and behavior patterns extraction but does not rely on purely statistical methods. Therefore, it not only can catch suddenly occurring outliers but also can reveal changes in the distribution of very noisy data.
Read more

We provide flexible engagement options to design and build an automated anomaly detection solution for your company. Contact us today to start with a workshop, discovery, or proof-of-concept (POC).