How American Airlines Charted a Course to Governance Excellence with Alation and Databricks

American Airlines is one of the world's largest air carriers, employing over 130,000 people globally and serving millions of customers across an extensive network. In recognition of its excellence, the business recently won several 2025 Friday Awards, including best airline redemption ability, best airline customer service, and best loyalty credit card. 

Behind this operational excellence lies a complex data ecosystem that captures everything from customer interactions to operational metrics, revenue management, baggage handling, and cargo operations. At a recent presentation at Databricks Summit 2025, their data leadership team shared a story revealing how they continue to modernize this data estate to deliver first-class customer service.

"American employs [is]... a global airline,” explains Anuradha Maradapu, Manager of Data Governance & Engineering at American Airlines. “So you can imagine we host huge amounts of data within the company." The airline's recent cloud migration has enabled unprecedented scalability and faster access to data, but it has also created exponential data growth that demanded a robust governance framework.

"Airline data is very complex,” shares Maradapu. This complexity, combined with the need for speed and self-service capabilities in a modernized environment, presented significant challenges that ultimately led American Airlines to implement an integrated data governance solution using Databricks Unity Catalog and Alation's Data Intelligence Platform.

American Airlines’ key data challenges

American Airlines faced several interconnected challenges that stemmed from its rapid data growth and complex operational requirements:

Data discovery and access issues. The airline struggled with making data easily discoverable across the organization. Speed to data access had created multiple redundant data sources, making it difficult for users to find the right information when they needed it.

Lack of business context. Technical data assets and field names were missing crucial connections to business terms and definitions. Even when users found data through tribal knowledge or by tracking down subject matter experts, the technical metadata lacked business context that would make it meaningful to non-technical stakeholders. What’s more, multiple definitions of the same metrics existed across different departments, leading to unclear data definitions and confusion about which version was authoritative. This created significant challenges for cross-functional collaboration and decision-making.

Reporting inconsistencies. The combination of unclear definitions, multiple data sources, and lack of standardization resulted in inconsistent reporting across the organization. Different teams could analyze the same underlying data and reach different conclusions based on varying interpretations.

High data management costs. The inefficiencies in data discovery, the need for manual processes to understand data context, and the time spent resolving inconsistencies all contributed to elevated costs in data management operations.

Limited data lineage visibility. Users couldn't easily trace data origins or understand how data moved through various systems, making it difficult to assess data quality and troubleshoot issues when they arose.

These challenges collectively created what Maradapu describes as "data gaps" – a fundamental disconnect between the technical reality of data assets and the business need for accessible, trustworthy, and contextual information.

As she explained, the rapid proliferation of data sources resulted in redundancy, poor discoverability, and fragmented knowledge. Even when users managed to locate data through tribal knowledge, the lack of linkage between technical metadata and business definitions made it difficult to derive value.

Objectives for data governance 

Recognizing that "data governance is a practice... implemented by people using accurate information through governed processes enabled by technology," American Airlines developed a comprehensive vision for data governance. Their approach was to embed governance directly within their data strategy rather than treating it as a separate initiative.

The airline established a clear data governance vision: "to embed the core data disciplines of data governance in processes with a mindset to foster a data-fueled culture that allows us to maximize the value of our data." This vision translated into three primary objectives that would guide their implementation strategy.

The first objective focused on increasing data availability by gaining efficiencies in data processes, increasing governance adoption, and comprehensively documenting available data assets. 

The second objective aimed to drive data trust by empowering domain experts with the right tools and processes, improving data awareness across the organization, and ensuring consistency in data understanding and usage. 

The final, third objective centered on reducing time and cost through standardized data policies and procedures, minimizing rework for data stewards, and establishing clear data accountability.

"Our whole goal was maximizing the value of our data to governance processes," Maradapu explains. These objectives were designed to create lasting value and support both current operational needs and future growth requirements.

Alation and Databricks implementation

American Airlines took a strategic approach to implementing its data governance solution, choosing to start with a focused pilot program before scaling across the enterprise. The team began by collaborating closely with business partners to introduce a scalable data governance program that could eventually be applied across all data communities within the organization.

The airline decided to focus initially on customer data as its minimum viable governance implementation. "We partnered with our customer data business partners and implemented this using them as our first minimum viable governance," Maradapu notes. This approach allowed them to test and refine their processes before expanding to other domains.

The technical foundation of their solution centered on the integration between Databricks Unity Catalog and Alation's Enterprise Data Catalog. American Airlines was already using Databricks to engineer and ingest data into its data lake, and migrating to Unity Catalog enhanced metadata capture capabilities. "The integration to Unity Catalog with Alation fueled our journey on data governance.”

The integration created several immediate benefits. Unity Catalog automatically generated technical metadata, which was then processed into Alation with minimal manual intervention. "With the Databricks Unity catalog, the technical metadata is being automatically generated; and [by] connecting this Databrick's Unity Catalog to our Alation, it just processed all the metadata into Alation. It created a lineage, [and] easy metadata extraction."

Working with their customer data team, American Airlines identified critical data elements and cataloged them comprehensively in Alation. This collaborative effort resulted in the creation of an MVP process for data curation that removed confusion around customer data availability and established clear pathways for deeper data inquiries.

A crucial aspect of their implementation was the launch of a data-steward community model. "We introduced a community model to empower our data stewards. We defined clear roles and responsibilities," Maradapu explains. This approach ensured that governance became a collaborative effort between IT and business partners rather than an IT-led initiative.

The integration also enabled comprehensive data lineage visualization, allowing users to discover data origins and track data movement throughout their systems. When data quality issues arose, the lineage capabilities combined with integrated data quality tools allowed teams to quickly assess upstream and downstream impacts.

To support adoption and reduce manual work, American Airlines implemented several automation features. They used robotic process automation to migrate tribal knowledge from tools like OneNote into Alation, and leveraged Alation's AI-powered natural language search capabilities to make data discovery more intuitive.

Key results from the Alation + Databricks implementation

The implementation of Databricks Unity Catalog integrated with Alation's Enterprise Data Catalog delivered significant improvements across American Airlines' data governance program:

Enhanced data discovery and access. Users gained the ability to easily find and understand data assets through Alation's intuitive search capabilities. The AI-powered natural language search not only discovered relevant data but also prioritized the most popular and frequently used datasets, making data discovery more efficient for business users.

Comprehensive data lineage visibility. The integration provided end-to-end data lineage that allowed users to trace data origins and understand data flow throughout their systems. "Data lineage is now available to discover the origin of data. We could see the movement of the data, the data flows," enabling faster issue resolution and impact analysis.

Standardized data definitions. Data definitions across customer data became standardized and consistent, eliminating the confusion that previously existed with multiple definitions of the same metrics. This standardization provided a foundation for reliable cross-functional reporting and analysis.

Automated metadata management. The integration enabled automated collection of technical metadata from Databricks Unity Catalog, significantly reducing manual effort required from data stewards. "Our stewards now spend their time in curating the data, standardizing the data and creating and keeping it consistent," rather than on manual metadata capture, Maradapu notes.

Improved data quality visibility. Integration with data quality tools provided users with real-time visibility into data health through Alation's interface. Quality issues were flagged with warnings, preventing users from accessing unreliable data and enabling proactive communication with data stewards.

Enhanced collaboration. The platform facilitated what Maradapu calls "data therapy sessions" – forums where domain experts could discuss data, share knowledge, and collaborate on data-related challenges. This collaborative approach transformed governance from an IT-driven initiative into a business-partnered effort.

Cost optimization opportunities. Analytics capabilities enabled the identification of underutilized data assets, creating opportunities for cost savings through data lifecycle management. Teams could now identify datasets that weren't being used and make informed decisions about data retention.

Scalable governance framework. The success with customer data provided a proven model that could be replicated across other domains within American Airlines, creating a pathway for enterprise-wide data governance standardization.

The results demonstrate that American Airlines successfully achieved their core objectives of increasing data availability, driving data trust, and reducing time and costs associated with data management. "The value of metadata is in its usage. The more they use this metadata to drive insights, that increases the value of the metadata," Maradapu observes, highlighting how the implementation created a positive feedback loop of increased adoption and value.

Conclusion: A governance framework that scales sky-high 

American Airlines' implementation of integrated data governance using Databricks Unity Catalog and Alation represents a successful transformation from siloed, manual data management to a collaborative, automated, and scalable governance framework. The airline's strategic approach of starting with a focused pilot program and gradually expanding across domains provides a blueprint for other large enterprises facing similar data governance challenges.

Looking toward the future, American Airlines plans to continue scaling data governance across additional data domains within their organization. The foundation they've established with customer data serves as a proven model that can be replicated for operations data, revenue management, cargo data, and other critical business domains.

Maradapu shares three key learnings from their journey that will guide future expansion efforts. First, "bridge the data to business gap" by associating technical metadata with business context to increase data value and create collaborative environments. Second, "trust your domain experts" by empowering people closest to the data to define its meaning and quality, which builds trust and drives adoption. Third, keep in mind that “governance is a team sport” and leverage tools and automations to remove the redundant work so domain experts can focus on data insights, awareness, and education.

The airline also plans to expand its use of business glossary features and continue developing what they call "data therapy sessions.” These sessions will play a crucial role in fostering the data-fueled culture that American Airlines envisions.

As Maradapu concludes, "Every role from a data engineer to an analyst, to a steward, and all the line of business users, they all play a crucial role. So if the IT team, technology team can create space for them by automations and tools and technology, the governance will now become a team sport. And can be very successful within an organization."

This collaborative approach positions American Airlines to continue maximizing the value of its data assets while maintaining the trust and accessibility that modern data-driven enterprises require.

Curious to learn how a data catalog can help you supercharge governance? Book a demo with us today.