Products

BigAI Assistant

An AI assistant that answers from your own internal documents and always cites the source so anyone can verify it.

Time to deploy: 3–5 weeks
Deployment: On-premise · Private cloud

The problem: the knowledge exists but nobody can find it

A 500-person company typically holds thousands of internal documents spread across SharePoint, Drive, email and shared folders. So people ask a colleague instead, and the same question gets answered hundreds of times a month.

Assistant does not replace your documentation — it makes it findable.

Features

What the product does

Cited answers

Every answer carries document name and page; one click opens the original passage.

Per-document permissions

Only retrieves documents the asker is entitled to see, synced from AD/LDAP.

Connects to existing sources

SharePoint, Google Drive, Confluence, network shares, databases and internal sites.

Embedded where people work

Web widget, Microsoft Teams, Slack or API for your internal applications.

How it works

Processing flow

1 · Ingest Parse, chunk and index into an internal vector store

2 · Retrieve Find relevant passages within the user permission scope

3 · Generate Summarise strictly from retrieved passages

4 · Cite Attach sources and confidence

Integrations

Connects to the systems you already run

SharePointGoogle DriveConfluenceMicrosoft TeamsSlackActive DirectoryKeycloak

Plans

Deployment plans

Prices below are indicative. A formal quote depends on user count, data volume and infrastructure model.

Starter

From $1,800/ month

For a single department or a narrow pilot.

Up to 100 users
3 document sources
Cloud deployment
Business-hours support

Request a quote

Most chosen

Business

From $4,800/ month

For company-wide rollout.

Up to 1,000 users
Unlimited document sources
On-premise or private cloud
SSO and per-document permissions
24/7 support, 99.5% SLA

Book a demo

Enterprise

For groups, banks and public sector.

Unlimited users
Self-hosted language models
Domain fine-tuning
Security review and pentest
Dedicated support engineer

Talk to an expert

FAQ

Frequently asked questions

It only answers from passages actually retrieved and always shows the source. When nothing relevant is found it returns "not found in internal documents" rather than guessing. Confidence thresholds can be tightened per document set.

In the internal deployment mode, the language model, vector store and documents all stay inside your infrastructure and the system needs no internet access. If you opt for an external API, we document exactly what is sent and add masking for sensitive fields.