Skip to content

Mesh Indexer Zendesk

The Mesh Indexer Zendesk node enables users to index and synchronize data from Zendesk into the MinuteView Mesh Indexing Platform. This task supports full and incremental indexing of records, including advanced options like vector embedding and OCR extraction.


🧠 Purpose

This node is used to:

  • Extract tickets and related data from Zendesk
  • Enrich content using optional features like full text, embeddings, OCR, and thumbnails
  • Push the indexed results into an Elastic Search index
  • Optionally generate vector embeddings using OpenAI

⚙️ Configuration Options

The following fields are available in the node configuration modal:

General Settings

FieldTypeDescription
Select IndexPicklistThe name of the target index in Elastic Search.
Index ActionPicklistDetermines the scope of indexing:
Full Index
Update From Date
Update Single Record

Service Account References

FieldTypeDescription
ServiceAccount-OpenAITextID of the OpenAI service account (optional).
ServiceAccount-Elastic SearchTextID of the Elastic Search service account.
ServiceAccount-ZendeskTextID of the Zendesk service account.

Index Parameters

FieldTypeVisible WhenDescription
Document IDTextUpdate Single RecordID of the specific record to update.
Start DateDateTimeUpdate From DateFilters records modified after this date.
White List FoldersText ListAlwaysList of folder paths to include in indexing.
DomainTextAlwaysThe Zendesk subdomain to extract from.
Record TitleTextAlwaysFormat for the indexed record title. Use placeholders if needed.
Sub TitleTextAlwaysFormat for the record subtitle (optional).

Advanced Options

FieldTypeDescription
Include Text VectorizationCheckboxEnables OpenAI-based vector embedding for content.
Include Full Text ContentCheckboxStores the entire text content of the document.
Include ThumbnailCheckboxGenerates a thumbnail for visual data sources if applicable.
Include OCRCheckboxEnables OCR (Optical Character Recognition) on files.

✅ Execution Logic

When executed, this node:

  1. Validates the presence of required service accounts (Zendesk, Elastic Search, and optionally OpenAI).
  2. Reads the selected index name and action.
  3. Builds an ItemQuery object based on user-supplied parameters.
  4. Instantiates a ZdeskDataSource object with the configured Zendesk subdomain.
  5. Launches the ItemIndexer, processing all matching records from Zendesk into the target index.
  6. Reports progress and completion status.

📌 Notes

  • The "Include Text Vectorization" option requires a valid OpenAI service account.
  • The task will fail if any required service accounts are missing.
  • The "White List Folders" field is currently implemented but may not be relevant for all Zendesk data structures. (To be confirmed manually)

🛠️ Troubleshooting

SymptomPossible Cause
Task fails with Required Parameters not foundInvalid or missing "Index Action" value
Task fails with Required Service Accounts not foundOne or more service account IDs are missing or incorrect
No records indexedEnsure filters like "Start Date" and "Domain" are correctly set

  • MeshIndexerAcc
  • MeshIndexerAdskVault
  • MeshIndexerNetworkFiles
  • MeshIndexerSharepoint
  • MeshIndexerMonday

Tentech 2024