Platform/Knowledge

Ground every prompt. In your data.

Eight bundled connectors, scheduled re-ingest, source ACLs preserved. A single Knowledge Query node sits before any LLM step. Whatever the user can't see, the agent can't surface — without you wiring up a thing.

Sources

Eight first-class connectors. More via MCP.

The bundled set covers the corpus most enterprises actually want grounded — wikis, drives, the warehouse, and the public web. Anything else, you wire through an MCP server.

Wiki
Confluence
Spaces, pages, attachments. Per-space ingest. Inherits Confluence ACLs at the page level.
Wiki
Notion
Databases & pages. Block-aware chunking. Workspace tokens with read-only scope.
Drive
SharePoint
Sites, libraries, folder trees. Microsoft Graph deltas for incremental sync.
Drive
Google Drive
My Drive + shared drives. Native Docs/Sheets/Slides. Drive ACLs preserved.
Object
S3
Bucket + prefix. Policy-driven include/exclude. PDF, DOCX, MD, plain text out of the box.
DB
Postgres
Read-only role, custom SQL or whole-table. Row-level security travels with the chunk.
DW
Snowflake
Warehouse + role + schema. Scheduled views. Tagged-column redaction.
Web
Web crawl
Domain-scoped. robots.txt-respecting. Sitemap-aware. Re-crawl on schedule.
Pipeline

From source to ranked chunk.

Each source flows through the same five-stage pipeline. You configure the source; the pipeline does the rest. Rebuild full or incremental, on a schedule or on demand.

01 — FETCH

Auth, paginate, delta

Vendor-specific connector pulls only what changed since the last sync.

02 — PARSE

Format-aware

PDF, DOCX, HTML, MD, structured rows. OCR fallback for scanned PDFs.

03 — CHUNK

Semantic + structural

Headings, sections, table rows. Chunk size tuned per source type, not globally.

04 — EMBED

BYO model

Bedrock, OpenAI, or your own gateway. Re-embeds on model upgrade — never silently.

05 — INDEX

ACL-tagged

Stored alongside the source's permission tuple. Query-time filter happens before retrieval.

The hard part

ACLs aren't an afterthought.

SECURITY-FIRST RAG

The user's permission tuple is in the query, not in a post-filter.

Most RAG systems retrieve first and filter after. That leaks: timing channels, ranking signal, "we know it exists, you just can't read it." Nexus stores the source's ACL alongside every chunk. The user's identity flows through the agent into the retrieval call, and the index filters before similarity search runs. If your user can't read the Confluence page, the chunk is never a candidate. Not redacted. Not sorted away. Never seen.

Wire your sources

30 minutes. Your real Confluence. Live.

We'll connect one of your sources during the demo, run an ingest, and show you a Knowledge Query node grounding a Claude response with citations.