The Customer

EDB (EnterpriseDB) is a global leader in Postgres software, services, and support. The company extends the power of open-source Postgres to meet enterprise demands—delivering high-performance database and analytics solutions that serve as the backbone for mission-critical workloads.

Within EDB, the Data Analytics and AI team plays a central role: providing business intelligence and supporting machine learning and AI innovation across sales, marketing, product, and engineering. The goal—like Postgres itself—is to make data consistent, reliable, and powerful for every stakeholder.

The Challenge

Before adopting Astronomer, EDB's data ecosystem was fragmented. Each department operated its own orchestration stack—GitHub Actions here, cron jobs there, and even Luigi scripts for more complex data engineering workflows. The result: siloed data processes, inconsistent reporting, and no single source of truth.

The EDB data team set out to unify orchestration across the company: a single platform capable of supporting data ingestion, transformation, analytics, and MLOps workloads—while scaling to future AI use cases.

The Solution

EDB selected Astronomer's managed Apache Airflow platform, Astro, for its flexibility, scalability, and ease of use. The team evaluated multiple options—including AWS Managed Workflows for Apache Airflow (MWAA) and Google Cloud Composer (GCC)—but found Astronomer uniquely suited to support diverse workloads while staying closely aligned with the open-source Airflow community.

EDB initially deployed Astronomer in a hybrid model, maintaining control of its data plane while Astronomer managed the control plane. Within a year, the company transitioned fully to Astro, offloading infrastructure management entirely and allowing the data team to focus on what mattered most—data reliability and innovation.

With Airflow now orchestrating data ingestion from CRMs, financial systems, and marketing platforms into a unified lakehouse, EDB established a single framework to serve analytics, data science, and AI workflows.

EDB uses Astro to orchestrate its entire machine learning lifecycle—from data preparation to model training and inference. Before Astronomer, most ML experimentation and training was done manually within notebooks, requiring engineers to kick off runs, monitor progress, and manage versioning by hand. With Astronomer-managed Airflow, these processes are now fully automated and repeatable.

This automation has enabled EDB to scale model training across multiple use cases and integrate those models directly into production analytics workflows—bridging the gap between data science and business decision-making.

The Results

Since adopting Astronomer, EDB has achieved a new level of collaboration and consistency across its data operations. The company's departments now share a single, scalable orchestration layer that powers everything from business dashboards to AI model training.

• One platform, all teams: Airflow unites three major teams—Data Analytics, Engineering, and Product—under a common orchestration framework.
• Single source of truth: Business metrics and KPIs are now standardized across departments, giving leadership a consistent, trusted view of company performance.
• Faster time to production: Analysts can launch new data products faster thanks to automation and streamlined workflows.
• Improved data quality: With centralized orchestration and data quality checks in place, EDB ensures cleaner, more accurate data enterprise-wide.
• Increased stakeholder trust: Leadership confidence in data-driven insights has risen as reliability and transparency have improved.

What's Next

EDB's next phase is focused on advancing MLOps and AI-driven workflows. With Astronomer powering the orchestration layer, the data team plans to expand automation, implement templated reusable workflows across departments, and further integrate AI capabilities. These AIOps efforts include automating model retraining, orchestrating LLM workloads, and embedding predictive analytics into internal business systems, driving faster, data-informed decision-making.