Organizations have more customer data sources than ever—purchase history, campaign activity, interactions across all channels, demographics and countless more—which together can reveal new, 360-degree customer insights, including predicting future needs and wants. Doing so, however, requires massively scalable data ingestion and modeling—as well as lightning-fast data queries, analysis, mining and AI processing.

Katana Graph™ is a graph intelligence platform for actionable Customer 360 insights, capable of integrating massive disparate customer datasets for advanced analytics and graph AI for predictive analytics, at tremendous scale and speed.

Katana Graph consolidates and renders huge volumes and varieties of customer data into a massively scalable, distributed graph database; performs high-speed graph and AI computing for customer identity resolution and segmentation; and enables advanced reporting and analytics, including predicting next best actions for customers using graph neural networks (GNN) and AI algorithms.

The Katana Graph Intelligence platform interconnects customer data at terabytes of scale, while also executing graph algorithms (with native support for traditional and deep learning models).

Core Katana Graph capabilities essential for Customer 360-driven insights include:

Consolidated graph representation of data. Katana Graph renders Customer 360 data sources into its resilient graph database. Graphs enable flexible-schema design to integrate various data structures and machine learning using graph embeddings and graph neural networks.

Highly accurate identity resolution/entity normalization in less time and lower cost, through graph analytics (e.g., similarity scoring, connected components, louvain clustering) and graph embeddings (i.e., projecting a high-dimensional graph into low-dimensional embedding space and using similarity calculations such as Jaccard).

Next Best Action prediction using Graph Neural Networks (GNNs). Katana Graph enables you to predict whether a customer will purchase a new product or service through GNNs, which can be trained to learn feature embeddings from neighboring nodes and neighborhoods in the graph.

Expanded customer segmentation for targeted sales/marketing spend. Segment customer profiles using Louvain clustering algorithm, and graph embedding. 

Advanced portfolio reporting and analysis using OpenCypher queries, distributed across cluster for high performance analytics and data visualizations.

About Katana Graph

The Katana Graph intelligence platform interoperates with its own graph database system, providing a single platform for:

Graph Queries (contextual search)

Graph Analytics (path finding, centrality and community detection)

Graph Mining (pattern discovery)

Graph AI & Deep Machine Learning (deep learning and prediction)

at unrivaled speed, using massive graphs, scalable to multiple clusters in any cloud environment.

Katana Graph is ready to fulfill your most pressing Customer 360 needs, now as well as in the future, at unmatched scale and performance.