Skip to content

AI Workers That Think, Learn & Act

Train them on your data. Delegate real authority. Every action auditable.

Your AI Team

starts here
click to explore
24/7Always On
Memory
100%Auditable

How VEP works

An AI worker in VEP is three separate things: a private brain that remembers, an engine that thinks, and a set of permissions that decides what it may do. You choose each one, and you can change them without losing the others.

The brain

Every worker has its own memory, and it is yours

The brain is Semantic DNA: a private, versioned store of facts with their sources. It grows from your chats, documents, mail, tasks and meetings. A new fact strengthens or supersedes an old one instead of overwriting it, so you can always trace what the worker believed, when, and why.

  • Every answer is grounded in that memory and can name where a fact came from.
  • You can inspect, correct and teach it: hand it documents, connect Gmail, Asana, Calendar, or explain in chat.
  • Memory is isolated per company; nothing is shared across customers.
Read about the Twin Brain →
Where it runs

Our EU cloud, or your own hardware

Workers and their brains run in VEP's EU cloud, isolated per company. Enterprise customers can run the whole platform on their own servers, in their own office, with support from us, so the brain never leaves their premises.

  • GPU machines can join as compute nodes to run open models locally (beta, on request).
  • The same worker, the same memory, wherever it runs.
Reference architecture: your AI worker on one computer in your own office (NVIDIA DGX Spark) → Private hardware and compute nodes →
The engine

Pick which intelligence does the thinking

The brain is separate from the engine. Each worker runs on the engine you choose, connected with your own subscription: Claude or ChatGPT (Codex) today, with more providers on the way. Switch the engine per worker in one click; the memory stays.

  • Model and reasoning effort are set per worker from the live catalogue.
  • Local open models on your own GPU are in beta.
Engines, models and the API →
Authority and audit

It does what you allow, and you can replay everything

You decide what each worker may do on its own: which tools, which accounts, which spend. Anything outside that authority stops and asks you. Every action, decision and message is logged with its context and can be replayed.

  • Human gates for writes to your CRM, mail and tasks.
  • One incident, one clear message: failures are explained in plain words, never raw logs.
Why we built it this way →

Simple, Transparent Pricing

Start free. Scale as your AI team grows.

Free

$0/mo
  • 1 AI employee
  • 100 messages/month
  • Web channel
Get Started

Family

$19/mo
  • 2 AI employees
  • 1,000 messages/month
  • Web + Telegram
Start Trial

Starter

$49/mo
  • 3 AI employees
  • 2,000 messages/month
  • Web + Email
Start Trial
Most Popular

Growth

$149/mo
  • 7 AI employees
  • 10,000 messages/month
  • All channels incl. Voice
Start Trial

Business

$399/mo
  • 20 AI employees
  • 50,000 messages/month
  • All channels + Opus model
Start Trial

Enterprise

Custom
  • Unlimited employees
  • Unlimited messages
  • Dedicated model + On-premise
Contact Sales

Ready-Made AI Employees

Deploy in 4 minutes. Train on your data.

Sales Development Rep

Qualifies leads, books meetings, follows up with prospects.

3x more qualified meetings Deploy Now

Customer Support

Handles inquiries, troubleshoots, escalates when needed.

Instant replies, human escalation Deploy Now

Content Writer

Blog posts, social media, newsletters in your brand voice.

Drafts in your brand voice, daily Deploy Now

Family Assistant

Schedules, shopping, reminders, household coordination.

Save 5+ hours per week Deploy Now

Meeting Scheduler

Coordinates across time zones, finds optimal slots.

90% less scheduling emails Deploy Now

VEP is the Virtual Employee Platform — AI workers with infinite memory, delegated authority, and full audit trails for customer support, sales, research, and HR.