# Use Cases (/docs/development/use-cases)



In BullSequana AI, a `use case` is an application or solution deployed on top of the platform to solve a concrete business problem.

Examples include:

* service desk copilots
* phone desk support assistants
* text-to-speech applications
* document AI pipelines
* retrieval and agent-based applications

What A Use Case Builds On [#what-a-use-case-builds-on]

Use cases usually combine platform capabilities from several layers:

* `Runtime` for networking, security, storage, observability, and delivery
* `CoreAI` for AI APIs, model lifecycle, vectors, and telemetry
* `ProAI` when analytics, ingestion, or BI components are also required

Typical Building Blocks [#typical-building-blocks]

A use case may rely on:

* the CoreAI API for inference
* MinIO for files and artifacts
* PostgreSQL or ClickHouse for application data
* Milvus for retrieval scenarios
* Argo Workflows or Temporal for orchestration
* Argo CD for deployment and lifecycle management

Delivery Model [#delivery-model]

Use cases should generally be delivered as platform-managed applications, with `Argo CD` as the preferred deployment path.

Helm charts and Kubernetes manifests are still useful, but they are best treated as deployment assets that Argo CD reconciles rather than as the final operating model by themselves.

Current Example [#current-example]

The first documented example in this section is:

* [CoreAI Doc AI Pipelines](/docs/development/use-cases/coreai-doc-ai-pipelines)