projection-patterns

To build read models from event streams, you implement projections that subscribe to events and update materialized views. This skill provides Python templates for live, catchup, persistent, and inline projections to create optimized query views for your CQRS read side.

The best way to optimize query performance for CQRS read sides is to denormalize data into specific projections. This skill provides patterns to aggregate event stream data into materialized views, enabling quick retrieval without querying the event store directly.

Yes, this skill supports database integration with asyncpg for projections. It includes Python templates that utilize asyncpg alongside asyncio to efficiently manage and persist event-driven projections into your database.

Yes, you can use Elasticsearch for search indexing in event-driven architecture projections. This skill provides specific templates for indexing events into Elasticsearch, allowing you to maintain searchable read models derived from your event streams.

You should use live projections for real-time updates, catchup projections to rebuild historical state, and persistent projections to maintain continuous, fault-tolerant read models. This skill details each type to help you choose the right pattern for your event-sourced system.