Pilot program open — 3 months free, special offers forever
AI search for Slack & Confluence · Zero data egress

Ask your org.
Get only what
you're cleared to see.

BeBrief is a permission-aware AI search tool for Slack and Confluence. Ask questions across your entire knowledge base — every answer is cited, filtered by your access permissions, and generated by a local LLM. No data ever leaves your org.

After install: open the admin dashboard, connect Confluence, then @mention BeBrief or send it a DM. Sign in →

Slack + Confluence ACL-filtered answers Local LLM Audit logs Multi-tenant

BeBrief uses Google OAuth for admin dashboard sign-in only. We do not access your Google Drive, Gmail, or any other Google services — only your email address for authentication. · Privacy Policy

Works with

Slack
Confluence
GitHub soon
MS Teams soon

Features

Everything your team needs to query internal knowledge

No hallucinations. No permission bypass. Just precise, cited answers from your actual data.

✅ Live

Permission-aware search

Every query is filtered against your org's ACLs in real time. Users only see documents they're permitted to access — enforced at the vector search level, not the UI layer.

✅ Live

Cross-source correlation

BeBrief finds connections across Slack threads and Confluence pages in a single answer. The Venn-diagram at its core — overlapping knowledge, surfaced precisely.

✅ Live

Semantic search

Natural language queries against all indexed content. High-dimensional vector embeddings mean you find what your team meant, not just what they literally typed.

✅ Live

Zero data egress

BeBrief manages dedicated infrastructure per tenant. LLM inference is local — no OpenAI, no Anthropic, no cloud AI. Your documents are never processed by a third-party provider.

✅ Live

Slack-native bot

@mention BeBrief in any channel. Get cited answers in Slack Block Kit format. Restricted content is delivered ephemerally — marked with ✅ and only visible to the querying user.

✅ Live

Confluence integration

OAuth 2.0 connect with Atlassian. Space-level filtering, page-level restriction enforcement, auto-linked identities via email matching. Incremental sync with CQL.

✅ Live

Admin dashboard

Web-based admin UI at /admin. Manage users, connectors, seats, and tenant settings. Google and Microsoft OAuth for admin sign-in.

✅ Live

Audit logs

Immutable audit trail of all queries, permission checks, and admin actions. Essential for compliance and security review. Append-only, per-tenant.

✅ Live

Pilot program

3-month pilot with all features, seat limits, and automatic inactive user cleanup. No credit card required. Lifetime special pricing for pilot members.

⚙️ Coming soon

GitHub connector

Index repositories, issues, and pull request discussions. Connect code knowledge to your Confluence docs and Slack conversations.

🗺 Planned

MS Teams bot

Full feature parity with the Slack bot for Microsoft Teams environments. Same ACL model, same local LLM, same cited answers.

🗺 Planned

Agentic RAG

Query expansion, multi-hop retrieval, and self-directed document traversal. Move from single-shot answers to structured research across your knowledge base.

How it works

From install to
permission-aware answers
in minutes.

01

Connect

Connect your Slack workspace and Confluence instance in minutes. Space selection, incremental sync, and identity auto-linking are all handled automatically.

Admin UI at /admin

02

Sync

BeBrief indexes your content with permission metadata attached. Every document carries its ACL — Slack channel membership, Confluence page restrictions, or public access.

10-min periodic + pre-query sync

03

Query

Ask anything in Slack. BeBrief filters the vector search by your identity's ACL, generates an answer from permitted documents only, and delivers it with full source citations.

Ephemeral delivery for restricted content

Query pipeline

User query embed Permission check filter Vector search generate Local LLM cite Cited answer source refs ephemeral if restricted

Every answer is traceable

Factual by design — every response cites its sources

BeBrief doesn't just surface an answer — it shows exactly which documents it came from. Every response includes source references linking back to the original Slack thread or Confluence page, so your team can verify and dig deeper.

→ Slack #engineering · thread
→ Confluence: Architecture Decision Record
→ Slack #product-design · discussion

Security

Permissions are never
an afterthought.

Security is the product, not a feature. Every architectural decision was made with the access control model in mind.

ACL-filtered vector search

Permission enforcement happens at search time — not in the UI, not as a post-filter. Your ACL is a hard gate on every query result.

Ephemeral delivery for restricted content

Restricted answers are delivered as ephemeral Slack messages — visible only to the querying user, marked with ✅, and never stored in channel history.

Just-in-time permission resolution

Permissions are resolved at the moment of each query against live source data. Revocations take effect rapidly — no stale grants persist.

Multi-tenant isolation

Fully isolated per-tenant: separate databases, separate vector stores, separate object storage. Cross-tenant queries are architecturally impossible.

Immutable audit logs

Every query, permission decision, and admin action is recorded. Append-only. Essential for compliance and security review.

No third-party data processing

LLM inference runs on BeBrief-managed servers. No OpenAI, no Anthropic, no cloud AI provider ever sees your documents.

Fail-closed sensitivity check

When in doubt, BeBrief treats content as restricted. Unrestricted pages must explicitly be marked public to surface openly.

ACL format

// Public Confluence page
"acl": ["public", "confluence_space:ENG"]

// Private Slack channel
"acl": ["slack_channel:C0123ABCD"]

// Restricted Confluence page
"acl": [
  "confluence_user:5b109f2e-...",
  "confluence_group:engineering"
]

// User identity grants
user.acl = [
  "public",
  "user:[email protected]",
  "slack_channel:C0123ABCD",
  "confluence_space:ENG",
  "confluence_user:5b109f2e-..."
]

Infrastructure

We manage it.
No big corporation
touches your data.

BeBrief runs dedicated infrastructure for each tenant — built entirely on open source components with no proprietary cloud dependencies. Your documents, embeddings, and generated answers stay within BeBrief-managed servers.

Relational databaseVector storePermission cacheMessage queueObject storageLocal inference

Dedicated per-tenant isolation

Separate database, vector store, and object storage per tenant. Cross-tenant access is architecturally impossible.

Open source foundation

Every infrastructure component is open source. No proprietary cloud lock-in. Auditable by your security team.

No third-party AI processing

LLM inference runs on our own servers. No API calls to OpenAI, Anthropic, Google, or any AI cloud provider.

Bring Your Own Storage — coming soon

Point BeBrief at your own object storage or database. Full data residency control for orgs with strict compliance requirements.

Pilot Program

Join the pilot.
Shape the product.

3 months free. All features. First to get updates. Special offers forever — we don't forget our heroes.

Limited offer

Pilot

Free for 3 months

+ lifetime special pricing after

  • All features unlocked
  • Slack + Confluence connectors
  • Permission-aware search
  • Admin dashboard + audit logs
  • First access to new features
  • Lifetime special pricing
Join the Pilot — free for 3 months

No credit card required. No commitment.

Enterprise

Custom
  • Unlimited seats
  • Priority support + SLA
  • Custom deployment (on-prem / VPC)
  • SSO integration
  • Custom connector development
  • Dedicated onboarding
Contact us

Pilot spots are limited. Seat-based billing launches after the pilot program. See all plans →

Limited pilot spots available

Be among the first.
We don't forget our heroes.

Join the pilot program — 3 months free, all features, first access to updates, and special pricing forever. We're onboarding teams hands-on.