AI-native segment discovery for customer intelligence

Building an AI-Ready Customer Data Foundation

Delaware North consolidated its first-party customer data on Databricks, spanning digital engagement, transactions, loyalty, and operational systems. The next challenge was ensuring that this data wasn’t just centralized but AI-ready. Traditional segmentation workflows relied on one-off queries and static extracts, making it difficult to discover new opportunities while supporting repeatable operations across multiple use cases.

Unity Catalog became the foundation for this transformation. By organizing customer datasets, features, and derived assets into a governed, searchable catalog, Delaware North established a shared semantic layer aligned to guest visitations, engagements, promotions, and transactions. This ensured that analytics, data science, and AI applications were all working from a consistent, trusted view of customer data.

AI-Native Discovery with Lakebase, Unity Catalog, and Delta Sharing

Neuralift is a Databricks “Built-On” AI partner that uses Lakebase as its persistent system of record for AI-ready features, segmentation outputs, and model artifacts. All assets are registered and governed through Unity Catalog, ensuring discovered segments are traceable, versioned, and accessible across teams without duplication.

Delta Sharing also plays a critical role in accelerating collaboration for Neuralift. Delaware North securely shared its customer datasets with Neuralift via Delta Sharing, eliminating the need for outside integration and data movement. Neuralift, in turn, shared the discovered segments and metadata back through the same mechanism, making insights immediately available in the Lakehouse for numerous use cases and activations.

“Because everything was cataloged and shared natively, we could move from data access to discovery almost immediately,” noted Neuralift Chief Data Officer Mike Maloney. “There was no translation layer slowing things down.”

Continuous AI-Driven Lift at Scale

With governance anchored in Unity Catalog, persistence through Lakebase, and secure exchange via Delta Sharing, Delaware North moved from episodic segmentation to continuous AI-native discovery.

Neuralift now supports segment and strategy discovery for 25+ recurring marketing and analytics use cases each month, including reactivation, loyalty tier optimization, promotion analysis, competitive analysis, and visit frequency acceleration. Instead of building static campaign lists, teams activate KPI-aligned segments knowing ahead of time how to influence net revenue, frequency, offer response, and lifetime value.

Neuralift surfaced previously unobservable guest clusters defined by patterns across digital engagement, gameplay, transactions, loyalty, winnings, and visit history data spanning rolling 12 months of player data and over 100M data points.

Sample view of AI-powered segmentation; not actual customer data

Each discovered segment is evaluated directly against business KPIs, enabling teams to prioritize segments based on measurable revenue lift potential before activation, rather than activating based on traditional tiering or RFM-based assumptions.

Before this foundation, segmentation required manual SQL queries, static extracts, and one-off reporting. Today:

• Discovery cycles compress from weeks to hours
• 400+ quarterly hours shift from manual analysis to optimization
• Segments persist as governed AI assets instead of temporary lists
• Performance is measured through controlled testing and KPI benchmarking by use case

Player IDs and insights flow directly back into Databricks for activation across CRM, paid media, digital engagement, and on-property programs. Delaware North did not simply improve segmentation. It established a scalable, always-accessible layer of customer intelligence that continuously identifies and activates new opportunities to lift core business KPIs.