---
schema_version: "1"
id: 8fdbc83b87981778e3c513f1bad5bb75d0cc850d7309c245cb9e4a3b8ba2fa98
title: OSI Workflow
collection: admin
version: current
locale: en-us
portal_url: https://docs.salesforce.com/admin/current/en-us/analytics/264-0-0/c360_a_sl_osi_workflow.md
canonical_url: https://help.salesforce.com/s/articleView?id=analytics.c360_a_sl_osi_workflow.htm&release=264.0.0&type=5
source_site: help.salesforce.com
content_type: documentation
content_hash: f3acc096d31eaa2109e347f0d34068020633f5b431e433f15c91f22ee6bf7b6c
description: The OSI workflow uses open-source translators and customer-managed infrastructure.
products:
  - Data 360
release: "264"
---
# OSI Workflow

The OSI workflow uses open-source translators and customer-managed infrastructure.

Before importing an OSI model into Tableau Semantics, make sure that these requirements are met.

- The underlying data exists in Data360 (either ingested directly or accessed via zero-copy architecture).
- Object and field names in Data360 exactly align with the definitions referenced in your OSI file.

Import a third party semantic model using OSI

The workflow includes these steps:

1. Export your existing model to the OSI format.
    
    How you translate your model into the OSI format depends on where your model exists now (Snowflake, dbt, and so on). While each vendor may provide different tooling for the translation, all published translators can be found in the[OSI GitHub repository](https://github.com/open-semantic-interchange/OSI/tree/main/converters). The outcome of this step is a portable semantic model in the OSI format.
2. Add required metadata to the OSI file
    
    Some metadata is not yet fully standardized in OSI v0.1 and must be manually added to the OSI file through custom extensions. Update your OSI file using the below properties to match your Data360 configuration:
    
    - Data set source: Update the source value for each data set to match the object API name of the corresponding data object in Data 360.
    - Data space: Provide the relevant data space API name.
        
        Example: `"data": "{"dataspace": "default"}"`
    - Object type (DLO or DMO): For each dataset in OSI, specify whether the object is a DLO or DMO in Data 360.
        
        Example: `"data": "{"dataObjectType": "DMO"}"`
    - Field type metadata: For each field, add the field data type that matches the field type in Data 360.
        
        Example: `"data": "{"dataType": "Text"}"`
3. Translate your OSI model into JSON compatible with Tableau Semantics
    
    Now that you have an OSI file with all the necessary metadata, run the OSI-to-Salesforce translator (via CLI or custom code). This produces a semantic model json file.that Tableau Semantics can consume.
4. Import your semantic model into Tableau Semantics
    
    Use Tableau Connect APIs to port the translated semantic model in your Tableau Semantics environment. [SeeTableau Semantics Layer APIs](https://developer.salesforce.com/docs/data/semantic-layer/guide/connect-api-in-depth-examples.html#create-a-full-semantic-model) for more information. This creates the final asset that Tableau Next and Agentforce can consume.

After the model is available in Tableau Semantics, you can query and use the model the same way you use other Tableau semantic models.

![Note](https://sf-zdocs-cdn-prod.zoominsoftware.com/tdta-analytics-semantics-264-0-0-production-enus/f7c99ab1-3439-4722-b925-34cb47455d89/images/icon_note.png)

Note This workflow is the initial interoperability foundation. Expect it to evolve into more guided and automated experiences as the OSI specification matures.