Skip to main content

​
General

VirtualBrain is an AI platform built for consulting, strategy, and investment professionals. It lets you upload documents to secure Brains, query them with AI-powered Chats, and run automated workflows with Apps. Every answer is sourced and traceable.
Those are AI models. VirtualBrain is the platform around your use cases that turns a model into client-ready advisory work. Four differences matter most:
  1. Sourcing: every answer cites the specific documents and passages behind it, so you can verify before it reaches a client.
  2. Industry fit: built for the deliverables consulting, strategy, and investment teams actually produce, with workflows that encode your methodology.
  3. Enterprise security: EU-hosted, ISO 27001 certified and SOC 2 Type II compliant, and your data never trains models.
  4. Model-neutral: we integrate advances from across the AI field, so you benefit from the best model for each task without betting your firm on a single provider.
Building your firm’s work on a single model concentrates risk, in three ways VirtualBrain is designed to absorb:
  • Cost: token-billed agentic tools can run up unpredictable bills when an agent loops without tight scoping. Because we own the orchestration, your costs stay scoped and forecastable.
  • Operating burden: with one model you own change management, workflow building, and keeping pace with every release, which often means rebuilding as new models break existing workflows. We absorb that integration work and verify nothing breaks before it reaches production.
  • Lock-in: today’s best model may not stay best for your use cases. Staying model-neutral means each shift in the field is an upgrade for you rather than a migration.
VirtualBrain is model-neutral and folds in advances from across the field as they become relevant to your workflows. Everything runs inside the EU, with no training on your data and no retention. Modes that don’t touch your documents, such as brainstorming and web research, run on EU-hosted providers that are themselves ISO 27001 and SOC 2 Type II audited. You always get a capable model for the task without having to standardize on one provider.
Yes. VirtualBrain is ISO 27001 certified and SOC 2 Type II compliant. Your data is encrypted in transit and at rest, hosted in the EU, and never used to train AI models. See Security and Compliance for details.
VirtualBrain is human-in-the-loop by default. Every claim carries a precise citation you can open and check against the source, Apps can include review checkpoints that pause for a person to validate intermediate results, and the full chain from input to output is transparent. When a partner asks where a number came from, you have an answer, not just a prompt history.
Each document you add gets an identity card: it’s classified, typed, and indexed. That lets VirtualBrain filter deterministically to the relevant set of documents before retrieving from them, so answer quality doesn’t degrade as a Brain grows past a handful of files. Context screening extends the same foundation across hundreds or thousands of documents at once.

​
Brains and Documents

A Brain is a secure knowledge repository where you upload and organize documents. Think of it as a dedicated workspace for a specific engagement, project, or topic.
VirtualBrain supports:
  • PDF
  • Microsoft Word (.docx)
  • PowerPoint (.pptx)
  • Excel (.xlsx) and CSV
  • Plain text (.txt), Markdown (.md), and HTML
  • Images
  • Audio (mp3, m4a) and MP4 video
  • Outlook emails (.msg)
  • Web pages and YouTube videos, via Web Content
Files can be up to 250 MB each.
Each Brain supports up to 10,000 files by default with a maximum file size of 250 MB per file. Need more capacity? Contact us to raise the limit.
Yes. In the chat input bar, Choose brain(s) to query lets you select several Brains for the same question. This is useful when information spans different knowledge domains. See Asking Questions.
We recommend organizing by engagement or knowledge domain:
  • One Brain per client project
  • Separate Brains for internal knowledge vs. client work
  • Topic-based Brains for ongoing research areas
See Brains for best practices.

​
Chats and Queries

Try these fixes:
  1. Be more specific: “Top 3 revenue risks in Q3 financials” vs. “Summarize everything”
  2. Check document quality: Poor OCR on scanned PDFs affects results
  3. Narrow your Brain: Query a focused Brain rather than a massive document set
Every answer includes citations. Click any citation to see the exact source passage and original document. Always verify critical information before presenting to clients. See Sourced Answers.
Yes. Toggle Web on in the chat input bar and VirtualBrain searches the web alongside your documents, citing web sources with their original URLs. Leave it off to stay strictly within your documents.

​
Apps

Apps are workflow that automate multi-step tasks. Instead of asking individual questions, Apps run structured analyses (such as due diligence reviews or market sizing) with consistent methodology.
Yes. The no-code App Builder lets you create custom workflows by combining blocks (prompts) for document analysis, synthesis, formatting, and more. See Building Apps.
VirtualBrain includes 50+ pre-built templates spanning the deliverables professional teams produce most: data-room scanning, proposals, benchmarks, market research, factsheets, and deliverable QA. Each is a starting point you can replicate and edit to match your own methodology.

​
Integrations

Yes. VirtualBrain connects to Microsoft OneDrive, Microsoft SharePoint, Google Drive, and Dropbox, syncing the documents you choose while mirroring their source permissions. See Cloud Storage Connectors.
Yes. VirtualBrain supports single sign-on with Google and Microsoft work accounts, including Microsoft Entra ID tenant configuration. SSO is set up with the VirtualBrain team during deployment, see Security and Compliance.

​
Support

  • Documentation: These pages
  • Start a conversation: Click on the question mark on the bottom right
  • Deployment team: Reach out to your dedicated AI Deployment Manager
  • Standard: Business hours (9am-6pm CET), response within 24 hours
  • Enterprise: Extended hours with priority response
Have a question not covered here? Contact leo-paul@getvirtualbrain.com.