Can DualDB use your own embeddings or models?

Model flexibility matters because enterprise stacks rarely stay static. Organizations may already have preferred providers for compliance, cost, latency or regional deployment reasons.

A model-agnostic data layer avoids over-coupling the application to one AI vendor. That makes it easier to evolve retrieval, generation and ranking strategies over time without redesigning the full platform.

This is also important for experimentation. Teams can compare providers and configurations while preserving the governance and structured reasoning layer that makes the application trustworthy.