Docs
Search K

Appearance

Mesh Indexer Autodesk Vault ​

The Mesh Indexer Autodesk Vault node allows you to perform content indexing operations on Autodesk Vault data and push structured results into a search index, supporting enriched content like full text, thumbnails, and embeddings.

This node is typically used in data processing pipelines where content from Autodesk Vault needs to be searchable and enriched with AI-generated vectorization or OCR.

🧠 Description ​

This node indexes selected documents from Autodesk Vault and sends them to a search index (e.g., Elasticsearch) for retrieval and analysis. It can optionally include full document text, OCR output, thumbnails, and text embeddings.

The indexing behavior is driven by a selectable action (full index, date-based updates, or single record), and supports content filtering based on folders and lifecycle states.

⚙️ Inputs ​

LabelTypeRequiredDescription
Select Indextext✅Name of the target index (e.g., vault-documents).
Index Actionpicklist✅Determines which indexing strategy to use:
• Full Index
• Update From Date
• Update Single Record.
Include Text Vectorizationcheckbox❌If checked, enables OpenAI-based embedding generation.
Include Full Text Contentcheckbox❌If checked, includes full text from documents.
Include Thumbnailcheckbox❌If checked, includes preview thumbnails.
Include OCRcheckbox❌If checked, runs OCR on document content.
Document IDtext❌Used only for Update Single Record. The Vault document ID to reindex.
Start DatedateTime❌Used only for Update From Date. Documents modified after this date will be reindexed.
White List Folderstext[]❌List of folder paths to include in indexing (optional filter).
Lifecycle Statestext[]❌Filters items by Vault lifecycle states.
Record Titletext❌Optional formatting string for record title metadata.
Sub Titletext❌Optional formatting string for record subtitle metadata.
ServiceAccount-Autodesk VaultserviceAccount✅The Autodesk Vault service account credentials.
ServiceAccount-Elastic SearchserviceAccount✅The search index service (e.g., Elasticsearch) credentials.
ServiceAccount-OpenAIserviceAccount✅The OpenAI service account used for embeddings.

🧪 Outputs ​

This node does not generate a specific output variable but updates the status of the job and reports indexing success or failure.

🔄 Behavior ​

Depending on the selected Index Action, the following behavior applies:

  • Full Index: All items in the whitelist and matching lifecycle states are indexed.
  • Update From Date: Only items modified after the specified Start Date are indexed.
  • Update Single Record: A specific Vault document is reindexed based on Document ID.

The node leverages multiple backend services:

  • VaultDataSource to connect to Autodesk Vault.
  • ElasticSearch for storing index data.
  • OpenAI to compute vector embeddings (if enabled).
  • OCR and thumbnail generation features if specified.

🧩 Example Use Cases ​

  • Index all released design documentation in Vault with searchable metadata and AI embeddings.
  • Update search index for items modified in the last 7 days.
  • Reindex a single corrupted or missed Vault item with complete metadata and thumbnail.

⚠️ Notes ​

  • The node requires three service accounts: OpenAI, Elastic Search, and Autodesk Vault.
  • For Update From Date and Update Single Record, only a subset of fields is evaluated.
  • Performance and index size may vary significantly depending on whether thumbnails, full text, and embeddings are included.
  • The index name must already exist in the search backend or be supported by your indexing infrastructure.

Tentech 2024