Data fabric, data lakehouse, and data mesh have recently appeared as viable alternatives to the modern data warehouse.
At Data Summit 2026, James Serra, data and AI Architect at Microsoft, held his session, “Choosing Between a Modern Data Warehouse, Data Fabric, Data Lakehouse, & Data Mesh,” providing a guided tour of each architecture to examine common data architecture concepts.
The annual Data Summit conference returned to Boston, May 6-7, 2026, with pre-conference workshops on May 5.
“It’s amazing to see some of these technologies come back around,” Serra said.
Serra examined how data lakehouses can help you achieve, how to distinguish data mesh hype from reality, and how to determine the most appropriate data architecture for your needs.
A relational data warehouse (RDW) is where you store data from multiple data sources into relational storage to be used for historical and trend analysis reporting to make better business decisions by getting greater insights into your company.
“It’s that single version of the truth so you have the same answer to each question,” Serra said.
A data lake is a schema-on-read storage repository that holds a vast amount of raw data in its native format until it is needed. “It’s a glorified file folder—just storage, where a RDW is both storage and compute,” he explained.
A modern data warehouse combines both a relational data warehouse and a data lake, he noted. There are five steps to building this architecture that includes data sources, ingestion, storage, transformation, modeling, and visualization.
“This concept is kind of going away but to ease the move to the cloud, some companies are still using this,” Serra said.
Serra defines a data fabric as a term used to describe the architecture of taking disparate systems and weaving them together, like fabric, to create a consistent layer on top of an organization’s data.
Data Fabric adds the following to a modern data warehouse:
- Data access policies
- Metadata catalog
- Master Data Management (MDM)
- Data virtualization
- Real-time processing
- APIs for retrieving metadata/data
- Building blocks/Services
- Products
A data lakehouse was established in 2020 and is the most popular data architecture right now, he said. This is a combination of a data lake and data warehouse. This is made possible by a Delta Lake, which is a transactional storage software layer that runs on top of an existing data lake, adding RDW-like features.
A data mesh is a decentralized approach to managing data, where multiple teams within a company are responsible for their own data, promoting collaboration and flexibility.
By implementing data mesh principles, the quality and accuracy of data can be enhanced, resulting in increased trust among businesses to utilize data more extensively for informed decision-making.
“It sounds great in theory but there’s a lot of problems with data mesh,” Serra said. “It’s good to learn and understand because you may take pieces of that for your solution.”
A data mesh enables different domains to create analytical data. It’s decentralized but it can be used in conjunction with other data architectures.
“It’s a concept, not a technology,” Serra said. “It requires a huge organizational and cultural shift.”
Many Data Summit 2026 presentations are available for review at https://www.dbta.com/datasummit/2026/presentations.aspx.