What makes CordonData different from other enterprise AI search tools?

CordonData combines on-premise RAG, document-level permission enforcement, automated PII/NHI/secret redaction, and advanced OCR in a single self-hosted package. Your data never leaves your infrastructure.

Can CordonData run completely air-gapped?

Yes. CordonData is designed for air-gapped, offline deployments. Run the entire stack — OCR, embedding, vector search, LLM inference, and SSO — entirely within your secure network.

How does document-level security work in CordonData?

ACL metadata is stored alongside vector embeddings. At query time, user identity is cross-referenced against this metadata, dynamically filtering the vector search space so users only see documents they have permission to access.

What document formats and languages does CordonData support?

CordonData supports PDF, DOCX, PPTX, XLSX, PNG, JPEG, TIFF, HTML, Markdown, plain text, and email formats. OCR supports 100+ languages including CJK, Arabic, Cyrillic, and Indic scripts.

Connectors & Integrations

Connect Every Source

Native connectors for SharePoint, Alfresco CMIS, Alfresco REST, and generic REST APIs. Sync documents, maintain ACLs, and index for AI.

Back to Home

// CONNECTORS

Connect to everything your team uses

CordonData connects to your existing infrastructure through native protocol connectors. No data migration, no file duplication — just secure, in-place indexing with full ACL sync.

connector protocols

14+

named enterprise systems

auth method types

100%

ACL sync on all connectors

CMIS Connector

cmis

Content Management Interoperability Services — the universal ECM protocol

Connects to any CMIS-compliant enterprise content repository. Documents stay in place; CordonData crawls, extracts ACLs, and indexes for AI — without touching your source system structure.

Supported Systems

Alfresco SharePoint FileNet Nuxeo OpenText Any CMIS repository

Bindings: Atom · Browser · WebServices

ACL extraction with group API resolution

Configurable CMIS query language filters

Max file size control · connection testing

REST API Connector

rest_api

Universal REST adapter with JSONPath field mapping and OAuth 2.0

Connects to any REST API source using configurable JSONPath field mapping. Pre-configured profiles for major platforms; fully generic for custom sources. Pagination modes: offset, page number, cursor, link-rel.

Works with

Confluence Jira Notion SharePoint Online Alfresco REST WordPress Drupal S3 Any REST API

Auth: None · Basic · Bearer · API Key · OAuth 2.0

SharePoint Online via Microsoft Entra (Azure AD)

S3 IAM permission resolution (bucket policy + SCP)

JSONPath ACL mapping · custom headers

Email Connector

Index mailboxes, folders, and attachments via IMAP. Supports delegated access and per-folder scoping.

IMAP · IMAPS · STARTTLS

Delegated + impersonation auth

Attachments · date/sender filters

Filesystem / NFS

Crawl mounted filesystems and NFS shares with POSIX ACL extraction and glob-pattern filtering.

Local · NFS mount

POSIX ACL extraction

Glob patterns · hidden file control

SMB / CIFS

Connect Windows file shares using domain authentication with full NTFS ACL extraction.

Windows domain auth

NTFS ACL extraction

Path-scoped · glob filtering

GitHub Connector

Index code, documentation, issues, pull requests, and wikis from GitHub or GitHub Enterprise.

GitHub · GitHub Enterprise

Code · issues · PRs · wikis

Fine-grained PAT · per-branch

What every connector includes

All connectors share the same production-grade indexing and security pipeline.

Incremental sync

Only changed content is re-crawled. Efficient for repositories with millions of documents.

Live ACL sync

Permissions from the source system propagate to CordonData. Access changes take effect on next sync.

Health monitoring

Sync health, failed documents, and last-crawl timestamps visible in the admin dashboard.

Microservice architecture

Each connector self-registers on startup via a standard API. Add new connectors without touching the core platform.

Under the Hood

What happens when you connect a source

From the moment you add a connector to the moment your team can ask questions — here's what CordonData does automatically.

1. Connect

You provide credentials. CordonData negotiates the connection — no document migration needed.

2. ACL Discovery

Permissions are read from the source and mapped to your identity provider — continuously, not just once.

5. AI Ready

Documents are searchable via hybrid retrieval — semantic meaning plus keyword precision — with full permission enforcement.

3. Content Processing

Text extracted. Scanned for PII, NHI, and secrets. OCR applied where needed. Metadata indexed.

4. Indexing

Content is indexed for both semantic and keyword retrieval. Embeddings stored alongside ACL metadata.

Incremental sync

Only changed documents are reprocessed. Large repositories sync efficiently without full re-indexing.

ACL follows the document

If permissions change in your source system, the change propagates to CordonData within the next sync cycle.

Sync health monitoring

Failed documents, skipped files, and sync health are visible in the admin dashboard in real time.