Open Source  ·  Self-hosted  ·  GDPR-ready  ·  Any LLM

One Brain for your team and for your AI.
AI answers you can trust.

veilSense gives your team a private knowledge hub — collect everything in one place, let AI assistants search it without hallucinating, and run it on your own infrastructure with any LLM you choose.

Request Demo
veilsense.company.local / brain
veilSense app
Why veilSense

The benefits that actually matter.

Built for teams that need AI to work on their terms — private, reliable, and fully in their control.

One place for everything your team knows
Collect chat responses, uploaded documents, research, meeting notes, and work results into a single searchable Brain. One source of truth — always up to date.
AI that answers from your data, not guesses
Assistants search your indexed Brain before responding. No hallucinations — every answer is grounded in content you control, not a language model's approximation.
Your LLM, your choice — including local models
Connect OpenAI, Claude, Gemini, or a local Ollama instance. Switch models per workspace without rebuilding anything. No vendor lock-in.
Live in minutes, not months
Pre-built assistants for Q&A, CV matching, and data modelling — ready to run from day one. No custom prompt engineering required to get started.
Open Source — just install with Docker
veilSense is fully open source. No license required — spin it up with a single Docker Compose command. The only costs are your own LLM API tokens and Tavily credits, billed directly by the provider.
Separate knowledge, restrict access
Organize teams, departments, or clients into isolated workspaces. Each workspace sees only what you allow — no cross-contamination, no accidental leaks.
Reuse your best prompts across the team
Build prompt templates once and share them within a workspace. Standardize how your team interacts with AI — consistent output, no duplicated effort.
Keep AI costs visible and under control
Set a monthly LLM budget per workspace. A live indicator shows current spend vs. limit so teams never exceed what you've reserved — no billing surprises.
Reach beyond your Brain when needed
When internal knowledge isn't enough, veilSense can extend search to the web via Tavily. Fresh external results are stored back into Brain automatically — filling gaps without leaving the platform.
Platform

See how it works.

Explore the core capabilities that make these benefits possible.

Collect everything.
Find anything instantly.

Brain is your team's central knowledge hub. Capture chat responses, documents, research, and results in one place — then let AI assistants search it so answers always come from your real content, never from guesswork.

  • ✓
    Ingest documents, chat results, and manual entries into one hub
  • ✓
    Semantic search — find content by meaning, not just keywords
  • ✓
    Knowledge graph shows how entries connect to each other
  • ✓
    Graph RAG ensures assistants retrieve only the most relevant context
Brain Wiki

Start immediately.
No prompt engineering needed.

veilSense ships with pre-built assistants you can point at your workspace and run right away. No custom LLM setup, no weeks of fine-tuning — just pick an assistant and go.

  • ✓
    Brain Q&A — ask questions, get answers grounded in your Brain
  • ✓
    CV Matcher — match candidates against your internal job requirements
  • ✓
    Data Model Assistant — inject schema context into any prompt automatically
AI Assistants

Keep sensitive data out
of the LLM — automatically.

veilSense detects and strips PII before any prompt reaches a language model. Workspaces are fully isolated so the right people see only the right knowledge — with no extra configuration needed.

  • ✓
    Auto PII masking — names, emails, phone numbers stripped before every LLM call
  • ✓
    Workspace isolation — each team's knowledge is strictly separated
  • ✓
    Encrypted storage — results and chat threads encrypted at rest
Privacy Shield

Reusable prompts.
Any LLM. Full history.

Build prompt templates once and reuse them across your workspace. Run them against any LLM — including local Ollama models. Every conversation is stored, resumable, and encrypted.

  • ✓
    Prompt templates — write once, reuse across the whole workspace
  • ✓
    Any LLM — OpenAI, Claude, Gemini, or a local Ollama model
  • ✓
    Persistent threads — full history, resumable context, encrypted at rest
  • ✓
    Ad-hoc chat — quick answers without setting up a thread
AI Chat
How it works

Up and running in three steps.

Deploy in an afternoon. Start getting value the same day — no AI expertise required.

1

Build your Brain

Add documents, write entries, and import existing content. veilSense indexes everything automatically — your team's knowledge becomes instantly searchable.

2

Pick your LLM and assistants

Connect any model — OpenAI, Claude, or a local Ollama instance. Enable the pre-built assistants that fit your team's workflows and start using them immediately.

3

Get reliable answers — privately

Ask questions and run assistants. Every response is grounded in your indexed Brain content. PII is masked automatically, and nothing leaves your server.

Your knowledge. Your Privacy.
Your infrastructure.

AI that works entirely on your terms — self-hosted, open source, and free to run. No vendor, no lock-in, no data leaving your walls. Just your team, your knowledge, and an AI that finally answers from truth.

Get in Touch