Optimize models for large datasets

**Description**
Currently handlers lacks proper serialization support.

**Proposed solution**
Add configurable serialization with sensible defaults.

**Priority**: Low/Medium

Start by locating the handlers and model paths, then trace how large datasets are currently processed and serialized. Define the configurable serialization options and sensible defaults from the existing behavior. Done means handlers support configurable serialization without losing support for large datasets.

Written by the indexing model from the issue text.