At MongoDB.local Milano 2025, I discussed how data modeling changes when applications, not databases, become the primary design constraint.

The session focused on practical modeling patterns in MongoDB, starting from real application requirements rather than abstract schemas. We explored how document models evolve from access patterns, domain boundaries, and workload characteristics, and why modeling decisions must be revisited as systems grow, integrate, and scale.

Rather than treating data models as static artifacts, the talk emphasized modeling as an iterative design process, tightly coupled with application behavior, performance expectations, and operational realities.

The examples presented were drawn from real production scenarios, highlighting common trade-offs around embedding versus referencing, data duplication, consistency boundaries, and schema evolution in long-lived systems.

The recording of the session is available below.