Infrastructure and Unity Catalog assets created by RAG Studio

Preview

This feature is in Private Preview. To try it, reach out to your Databricks contact.

Looking for a different RAG Studio doc? Go to the RAG documentation index

RAG Studio will create the following items in your configured Databricks Workspace and Unity Catalog schema.

Warning

Infrastructure will incur charges per your Databricks contract.

Note

Below, any value that appears in curly brackets {} is replaced by the value for that configuration setting in rag-config.yml. For example, {name} will be replaced by the RAG Application name that you configured.

Infrastructure created in the Databricks Workspace

Component

Default Value

Configurable Value

Created ONCE for the RAG Application

Vector Search Endpoint

{name}__vs_endpoint

global_config.vector_search_endpoint

MLflow Experiment name

/Shared/{name}__experiment/

global_config.mlflow_experiment_name

Created FOR EACH Version

Model Serving Endpoint to host the 🔗 Chain

rag_studio_{name}_{env_name}

not configurable

Created FOR EACH Environment

Databricks Jobs

[{name}][{env_name}] {job_name}

not configurable

Note

The MLflow experiment name must start with either /Shared or /Users.

Note

The Databricks Jobs will be replaced with Databricks managed serverless compute services in future versions.

Assets created in the Unity Catalog schema

Component

Default Value

Configurable Value

Created FOR EACH Version

Unity Catalog Delta Table with the output from the 🗃️ Data Processor

{name}__embedded_docs__{random_uuid}

data_processors[].destination_table.name (if specified, index will be called {provided_value}__{version_id})

Unity Catalog Vector Index of the Delta Table created by the 🗃️ Data Processor

{name}__embedded_docs_index__{random_uuid}

data_processors[].destination_vector_index.databricks_vector_search.index_name (if specified, index will be called {provided_value}__{version_id})

Unity Catalog Model for the 🔗 Chain

{name}__chain__{version_id}

not configurable

Inference Table for the 🔗 Chain’s Model Serving Endpoint

rag_studio_{name}_{env_name}_payload

not configurable

Created FOR EACH Environment (details

🗂️ Request Log for user traffic

rag_studio_{name}_{env_name}_assessment_log

not configurable

👍 Assessment & Evaluation Results Log

rag_studio_{name}_{env_name}_request_log

not configurable

Created FOR EACH 📖 Evaluation Set

👍 Assessment & Evaluation Results Log

{name_of_evaluation_set_table}_assessment_log

not configurable

🗂️ Request Log

{name_of_evaluation_set_table}_request_log

not configurable