Ellipse Gradient for Header

Glean Integration Guide

Connect Glean’s REST APIs for search, conversational queries, content indexing, and workplace data operations with enterprise workflows and applications.

Glean integration options at a glance

Glean’s primary integration model is REST API access for search, conversational queries, document indexing, People operations, metadata retrieval, and related workplace data capabilities. Martini can consume these APIs from workflows, authenticate with a Glean bearer token, transform payloads, apply business rules, and expose an application-specific REST API façade. Content synchronization can run on a schedule or be initiated by events from source applications. Glean supports indexing at platform scale, but exact bulk and asynchronous operations depend on the applicable API and tenant configuration. General-purpose webhooks, GraphQL, SOAP, direct database access, and a broad file-storage API were not confirmed.

Integration pointSupported by Glean?Common use casesHow Martini supports it
REST APIsYesGlean publicly documents REST APIs for search, chat or conversational queries, document indexing, People operations, metadata retrieval, and related platform operations.Martini can consume Glean REST APIs from workflows, map requests and responses, apply business rules, and expose controlled REST APIs for downstream applications.
Indexing APIsYesIndexing APIs can create or update Documents, withdraw indexed content, and submit source metadata, identifiers, and visibility information.Martini can extract changed source content, transform it to Glean’s indexing model, submit updates, and handle deletion, retry, and checkpoint logic.
Search APIsYesSearch APIs support submitting queries and returning structured Search results with metadata, snippets, relevance information, and source references.Martini can validate requests, call Glean, transform results, combine them with other systems, and return an application-specific response.
Conversational APIsYesGlean provides chat or conversational operations that can return Chat responses and references to source documents.Martini can orchestrate conversational requests, enforce request policies, map responses, and expose them through a controlled API façade.
Bulk / async / batch APIsLimitedGlean’s indexing model is intended for content ingestion at platform scale, but exact bulk, asynchronous, and batching operations depend on the applicable indexing API and tenant configuration.Martini can use bounded batches, checkpointing, concurrency controls, idempotent identifiers, and retry handling without assuming one request can process a complete repository.
Webhooks / outbound callbacksNot confirmedGeneral-purpose webhook coverage for document, People, search, or indexing changes was not established. Specific product or tenant callbacks require confirmation.Martini can receive events from source applications when available, or use scheduled incremental synchronization when Glean callbacks are not confirmed.
File / attachment APIsLimitedGlean indexes document content and metadata, but a general-purpose file-storage or attachment API was not confirmed.Martini can map content into supported indexing payloads or coordinate with a source connector, subject to Glean’s applicable content requirements.
AuthenticationYesGlean API access commonly uses an API token supplied as a bearer token in the HTTP Authorization header, with permissions managed by Glean administrators.Martini can externalize the tenant URL, token, timeout, and retry settings and store the token in secrets management rather than workflow literals.

How Glean exposes data and business events

Glean REST APIs

Glean’s primary documented integration mechanism is REST API access for search, chat, indexing, People operations, metadata retrieval, and other tenant-enabled workplace data operations.

Martini implementation pattern

Martini implementation pattern: a workflow or exposed Martini API receives a request, builds the Glean REST request, supplies the bearer token from secrets, validates the response, applies business rules, and returns or persists the transformed result.

Implementation sequence

Receive a request or start the workflow
Load the Glean base URL and bearer token from secure configuration
Build and send the Glean REST request
Validate the response and classify errors
Map the result to the target application model
Return the response or write the transformed data

Glean Indexing APIs

Glean indexing APIs load or update Documents and related content metadata from external systems. Required fields, content types, permission metadata, and lifecycle operations depend on the applicable indexing API and tenant configuration.

Martini implementation pattern

Martini implementation pattern: a scheduled or source-event workflow extracts changed content, maps source identifiers and visibility metadata, submits bounded indexing requests, and records checkpoints so retries do not create duplicate Documents.

Implementation sequence

Retrieve changed source content
Assign a deterministic source identifier
Map content, metadata, and visibility information
Submit a create or update request to Glean
Process deletions, withdrawals, or disabled content
Store the checkpoint and outcome

Glean Batch Processing

Glean’s indexing model is intended for high-volume content ingestion, but exact bulk, asynchronous, and batch endpoints should be confirmed for the target tenant and API version.

Martini implementation pattern

Martini implementation pattern: workflows divide source content into bounded batches, control concurrency, checkpoint successful progress, retry transient failures with backoff, and route permanent payload or authorization errors for review.

Implementation sequence

Partition changed content into bounded batches
Submit each batch within configured concurrency limits
Record successful identifiers and checkpoints
Retry transient failures with exponential backoff
Route invalid or unauthorized items for review
Reconcile source and Glean synchronization status

Common Glean integration patterns

Pattern 1: Synchronize enterprise content into Glean

When to use this pattern

Use this pattern when documents from systems such as Confluence, ServiceNow, Salesforce, Jira, or SharePoint must become searchable in Glean while preserving source identifiers, visibility metadata, and lifecycle state.

Integration direction
Source application
Martini
Glean
Example Mapping
Glean FieldCanonical FieldTarget Field
sourceIdexternalDocumentIddocumentId
titlecontentTitletitle
bodysearchableContentcontent
updatedAtsourceModifiedAtmodifiedTime
Martini implementation pattern

A Scheduler Trigger or source-system event starts a workflow that retrieves changed content, maps it to the Glean indexing model, applies permission and deletion rules, and submits idempotent updates. The workflow checkpoints progress, retries transient failures, and routes invalid indexing payloads to an exception path.

Martini capabilities used
  • workflows
  • API consumption
  • scheduling
  • data mapping
  • business rules
  • error handling

Pattern 2: Provide Glean search through an internal application

When to use this pattern

Use this pattern when an employee portal, intranet, or business application needs Glean search without exposing Glean credentials or coupling clients directly to Glean’s response schema.

Integration direction
Employee application
Martini
Glean
Example Mapping
Glean FieldCanonical FieldTarget Field
querysearchTextquery
filterssearchConstraintsrequest parameters
results[].documentsearchHititems[].source
results[].snippetresultSummaryitems[].summary
Martini implementation pattern

Martini exposes a controlled REST API that authenticates the caller, validates query and tenant parameters, invokes Glean search, transforms Search results, and applies response and audit rules. Authentication failures, rate limits, and Glean availability errors are classified separately from invalid client requests.

Martini capabilities used
  • API exposure
  • API consumption
  • data mapping
  • business rules
  • authentication
  • error handling

Pattern 3: Combine Glean answers with transactional data

When to use this pattern

Use this pattern when Glean provides discovery or conversational context but an authoritative system must supply current status, approval, balance, entitlement, or operational details.

Integration direction
User application
Martini
Glean
ServiceNow or Salesforce
Example Mapping
Glean FieldCanonical FieldTarget Field
queryuserQuestionGlean query
chatResponse.referencessourceReferencesresponse.sources
sourceReference.idauthoritativeObjectIdServiceNow or Salesforce identifier
transaction.statuscurrentStatusresponse.transactionStatus
Martini implementation pattern

The Martini workflow calls Glean for Search results or Chat responses, extracts trusted source references, retrieves current data from the authoritative application, and combines the results while distinguishing indexed knowledge from live transactional state. Missing references and downstream failures are handled without presenting stale data as current.

Martini capabilities used
  • workflow orchestration
  • API consumption
  • data mapping
  • conditional routing
  • business rules
  • error handling

Pattern 4: Synchronize People and organizational context

When to use this pattern

Use this pattern when the Glean tenant supports the required People or identity operations and approved employee attributes must be synchronized from Workday or another source system.

Integration direction
Workday
Martini
Glean
Example Mapping
Glean FieldCanonical FieldTarget Field
employeeIdpersonExternalIdPeople identifier
departmentorganizationUnitdepartment
locationworkLocationlocation
employmentStatuspersonStatusstatus
Martini implementation pattern

A scheduled workflow extracts permitted changes, normalizes identifiers and status values, validates tenant-specific requirements, and submits supported People updates. The workflow handles disabled users, rejected attributes, authorization failures, and checkpoint recovery conservatively because exact People operations depend on tenant configuration.

Martini capabilities used
  • scheduling
  • API consumption
  • data mapping
  • validation
  • business rules
  • error handling

Applications commonly integrated with Glean

Glean can be integrated with enterprise applications whose content, knowledge, or organizational data needs to be searchable and usable in workplace experiences. Martini can coordinate the application APIs, source permissions, indexing requests, and downstream responses without requiring a dedicated Martini connector.

Application Scenario Direction Martini Pattern
Salesforce Make Accounts, Contacts, Opportunities, and related knowledge discoverable through Glean, or enrich CRM-facing applications with Glean search. Salesforce → Martini → Glean A scheduled or source-triggered workflow retrieves changed Salesforce content, maps stable identifiers, metadata, and visibility information to Glean’s indexing model, and submits idempotent updates. A separate Martini API can call Glean search for CRM-facing applications.
ServiceNow Make knowledge articles, incidents, catalog information, and operational documentation searchable through Glean. ServiceNow → Martini → Glean Martini retrieves changed ServiceNow content or receives supported source events, validates and transforms the payload, and submits document updates or withdrawals to Glean. Search responses can also be combined with live ServiceNow status data.
Jira Index project issues, engineering work items, and related documentation for enterprise discovery. Jira → Martini → Glean A Martini workflow extracts changed Jira items, preserves source identifiers and permissions, maps content and metadata to Glean Documents, and retries transient indexing failures while routing invalid payloads for review.
Confluence Make technical documentation, policies, and team knowledge searchable in Glean. Confluence → Martini → Glean Martini schedules incremental Confluence extraction, maps page content and source references to Glean indexing requests, handles page deletion or archival, and stores checkpoints so failed runs can resume safely.
Slack Improve discovery of permitted conversations and shared knowledge, subject to source permissions and retention policies. Slack → Martini → Glean Where the relevant APIs and tenant policies permit it, Martini retrieves eligible Slack content, applies visibility rules, and submits bounded indexing batches to Glean using deterministic source identifiers.
Google Drive Index shared documents and folders so employees can find enterprise content through Glean. Google Drive → Martini → Glean Martini retrieves changed Drive content and metadata through supported source APIs, maps permissions and source references, and sends create, update, or withdrawal requests to Glean with checkpointing and retry controls.
Workday Synchronize employee or organizational context and make selected HR knowledge discoverable. Workday → Martini → Glean A scheduled Martini workflow extracts permitted employee or organizational changes, normalizes identifiers and status values, and submits supported People or indexing updates to Glean after tenant-level capability validation.
Microsoft SharePoint Index policies, departmental documents, and intranet content for unified discovery. Microsoft SharePoint → Martini → Glean Martini retrieves changed SharePoint content, maps source paths, metadata, permissions, and lifecycle state to Glean’s indexing model, then processes bounded batches with idempotent identifiers.

How to build a Glean integration in Martini

Objective

Configure the Glean tenant URL, bearer token, timeout, and retry settings without embedding secrets in workflow logic.

Instructions in Martini

  • Use Martini REST API consumption for Glean calls
  • Externalize the tenant or API base URL and operational settings
  • Store the Glean API token in Martini secrets management
  • Keep source-system credentials separate from Glean credentials

Objective

Select a trigger that matches the synchronization or API façade requirement rather than assuming Glean provides universal webhook events.

Instructions in Martini

  • Use an exposed Martini API for interactive search or chat requests
  • Use a Scheduler Trigger for incremental indexing
  • Receive source-application events when those applications provide them
  • Confirm any Glean-specific callback capability before relying on it

Objective

Read source content, People data, or client requests and preserve identifiers, timestamps, permissions, and source references.

Instructions in Martini

  • Retrieve only the changed source objects where incremental APIs are available
  • Handle documented pagination or continuation cursors
  • Capture deletion, archival, and permission changes
  • Use stable source identifiers for later reconciliation

Objective

Coordinate Glean calls with source applications, checkpoints, batching, conditional routes, and downstream operations.

Instructions in Martini

  • Build the workflow around REST API requests and response validation
  • Use bounded batches for large indexing transfers
  • Store checkpoints after successful units of work
  • Separate transient failures from permanent validation or authorization failures

Objective

Convert source payloads and Glean responses into canonical models that can be reused across applications.

Instructions in Martini

  • Map content, metadata, identifiers, and visibility information to indexing requests
  • Transform Search results and Chat responses for consuming applications
  • Preserve source links and references
  • Use reusable mappings for Glean request and response schemas

Objective

Enforce tenant, authorization, data visibility, lifecycle, and source-of-truth rules before writing or returning data.

Instructions in Martini

  • Validate caller and tenant parameters for exposed APIs
  • Apply source permission and visibility requirements
  • Route deleted or disabled content to withdrawal handling
  • Keep Glean discovery data distinct from authoritative transactional state

Common Glean data objects used in integrations

ObjectTypical UseCommon target systemsMartini handling
DocumentsSearchable content indexed by Glean, including source-system documents and other content objects.Glean, Confluence, ServiceNow, Salesforce, Jira, SharePoint, Google DriveMartini maps source identifiers, content, metadata, permissions, and lifecycle state, then submits idempotent create, update, or withdrawal requests through indexing APIs.
PeopleUser and employee profiles used for people search, personalization, and identity-aware experiences.Glean, Workday, identity systems, employee portalsMartini can transform permitted employee and organizational attributes and submit supported updates where the tenant API and permissions allow it.
CollectionsCurated groups of documents or content assembled for organizational use.Glean, intranets, employee portalsMartini can synchronize supported collection metadata or use collection information when shaping downstream search and content responses, subject to the applicable API.
DatasourcesSource systems or content repositories from which Glean indexes content.Glean, source application catalogs, administration toolsMartini can use datasource metadata to route synchronization logic and associate indexed content with its originating system when supported by the API.
Search resultsStructured results containing document metadata, snippets, relevance information, and source references.Employee portals, intranets, CRM applications, service portalsMartini validates and transforms results, preserves source references, applies response rules, and can combine discovery results with authoritative transactional data.
Chat responsesConversational responses that may include references to source documents.Employee assistants, portals, custom applicationsMartini calls conversational endpoints, applies authorization and response policies, maps the response schema, and returns it through a controlled API.

Authentication and security considerations

Bearer-token authentication

Glean API access commonly uses an API token supplied as a bearer token in the HTTP Authorization header. Token issuance and permissions are managed by Glean administrators.

Secrets and visibility

Store the Glean token, tenant URL, and related configuration in Martini secrets and environment-managed settings. Do not embed tokens in workflows, mappings, source control, logs, or error payloads.

Permission-aware design

Glean results may be permission-aware. Align the requesting user, API credential, indexed-content visibility, and source permissions. Do not use a broad service credential for user-specific search unless the Glean security design explicitly supports it.

Operational considerations for Glean integrations

Rate limits and pagination

Confirm tenant-specific limits and quotas. Use bounded concurrency, batch sizing, exponential backoff, and handling for HTTP 429 responses. Continue through documented pages, cursors, or continuation tokens rather than relying on a fixed page count.

Incremental synchronization

Prefer source timestamps, versions, change tokens, or source events. Store checkpoints and use deterministic identifiers so a failed run can resume without reprocessing an entire repository or creating duplicate Documents.

Lifecycle and schema changes

Handle updates, deletions, archived content, disabled users, permission changes, and replaced source objects. Isolate Glean mappings in reusable assets and monitor API version, field, enum, and indexing requirement changes.

Observability and testing

Log correlation identifiers, source identifiers, operation types, response status, and retry counts without logging tokens or sensitive content. Test representative indexing, search, visibility, error, and reconciliation scenarios before production deployment.

Why use Martini instead of scripts or point-to-point integrations?

Orchestrate more than an API call

Scripts can call Glean, but enterprise integrations also need source extraction, incremental checkpoints, batching, permission rules, data transformation, retries, and operational visibility. Martini provides workflows and reusable integration assets for these concerns.

Expose controlled APIs

Martini can provide an application-specific REST API façade over Glean search or conversational operations. This centralizes caller validation, tenant routing, response mapping, audit behavior, and error handling without exposing Glean credentials to every client.

Maintainable integration design

Martini separates vendor API consumption from business rules and target-system writes. This makes it easier to combine Glean discovery with authoritative Salesforce, ServiceNow, or other application data while preserving clear source-of-truth boundaries.

Frequently asked questions

How can Glean be integrated with enterprise systems?

Glean can be integrated through its documented REST APIs for search, conversational queries, document indexing, People operations, metadata retrieval, and related workplace data operations. Enterprise workflows can use scheduled incremental synchronization or source-system events, while a controlled API façade can expose Glean capabilities to internal applications.

Can Martini integrate with Glean?

Yes. Martini can consume Glean’s REST APIs from workflows, authenticate with a bearer API token stored in secrets management, map and transform requests and responses, submit indexing operations, and expose REST APIs that orchestrate Glean operations.

Do I need a connector to integrate Glean with Martini?

No. A dedicated Glean connector is not required. Martini can integrate with Glean using its confirmed native REST APIs and bearer-token authentication, together with source-application APIs, events, files, or other supported mechanisms where applicable.

Is there any extra Lonti cost to integrate Glean with Martini?

Lonti does not charge an additional per-connector or per-vendor fee to integrate Glean with Martini. Integrations are subject to the provisioned capacity of the Martini environment. Separate costs may apply from Glean, infrastructure providers, source applications, or other third-party systems based on their subscriptions, usage, and deployment models.

Which Glean integration methods should architects use?

REST APIs are the primary and recommended mechanism for new integrations. Use search and conversational APIs for discovery experiences, indexing APIs for content synchronization, and People operations where enabled by the tenant. Bulk or asynchronous processing should be confirmed for the applicable indexing API before production use.

Does Glean provide webhooks or event notifications?

General-purpose webhook coverage for all document, People, search, or indexing changes was not confirmed. Confirm callbacks for the specific Glean product or tenant. Otherwise, use scheduled incremental synchronization or events from the source application.

How does Martini synchronize and transform Glean content?

Martini can retrieve changed source content, map identifiers, content, metadata, permissions, and lifecycle state to Glean’s indexing model, and submit bounded, idempotent requests. Checkpoints, pagination handling, deletion processing, and retries help the workflow resume without duplicating content.

How are errors, retries, and duplicate Documents handled?

Martini can classify authentication, authorization, validation, rate-limit, availability, and source-system failures. Transient failures can use bounded retries and exponential backoff, while permanent failures can be routed for review. Deterministic source identifiers and idempotent indexing reduce duplicate Documents.

Related Martini documentation

APIs

Workflows

Data

Security

Operations

Connect Glean with Martini

Use Martini to build secure, maintainable Glean integrations for search, conversational experiences, content indexing, and enterprise data synchronization.

EXPLORE MARTINI
BOOK A DEMO