- Made in Germany
- Platform hosted in Germany
- GDPR-compliant
AI-VI. The AI with a human touch.AI-VI Cybersecurity Coach
Your personal AI avatarfor realistic training scenarios
AI-VI provides lifelike AI training environments in which AI avatars work with you interactively in real-life scenarios: practising workflows, assessing know-how or preparing for crisis situations such as cybersecurity incidents. AI-VI takes on different roles and uses tools like emails and video calls, yet draws exclusively on professionally approved content, without hallucinations, traceable, transparent and GDPR-compliant.
Two views. One session.
The moderator runs the fire drill while participants experience it in sync, each role with its own view.
Product tour
This is what working with it looks like.
Playbooks are built in the configurator with AI support. Live, the moderation steers phases, checkpoints and injects, followed by the debrief.
The film
13:17. Every system down.
Ninety seconds of the real thing: an attack of the kind that happens every day, a crisis team losing its grip. And one sentence at the end that turns it around.
Note: These are demo views using fictitious company data, shown for illustration only. The user interface is continuously evolving and may change as new features are added, so the exact content will differ.
The real thing
The first three minutes decide.
This is what the crisis team chat looks like while an attack is running. Six messages, three minutes, and nobody has the full picture.
That is exactly the sequence the Cybersecurity Coach rehearses, before it is real. With your systems, your roles and your procedures.
Crisis team
· 7 participants · active
CISOAll systems down. Crisis team. Now.13:14
Head of ITNo access at all. Servers encrypted.13:14
ManagementWhat is going on???13:15
HRNobody can get into the system.13:15
MarketingOur website is offline.13:16
FinanceAll payments blocked.13:16- UnknownYour files are encrypted. 72:00:00
Know-how and training
Preparing and running good, subject-specific training costs time and ties up resources.
Knowledge and experience often sit with a few employees who are tied up in day-to-day business.
Self-studies or training with purely generative AI carry risks when training results and AI answers cannot be verified or traced.
The solution AI-VI. The AI with a human touch.
No hallucinations
Answers are given only on the basis of validated content.
Know-how
Verified content, adaptable and interactive with real scenarios.
Data sovereignty
Platform hosted in Germany, optionally on-site. By default, avatar videos are generated by a US provider, covered by EU Standard Contractual Clauses.
Protected & multilingual
70+ avatars, 70+ languages.
Time savings
Access to existing AI trainings around the clock (24/7/365), no time-consuming organisation of meetings, equipment or rooms, remote participation possible.
Large enterprises use AI-VI for cybersecurity training
Trusted by companies and partners
Fire drills for IAM & encryption
For an internationally operating technology company, we ran several fire-drill simulations in the security-critical areas of identity & access management (IAM) and encryption. The goal: test the teams' ability to react under realistic conditions and sustainably raise the security level, playful rather than dry.
- 1
The solution
Realistic scenarios were designed together with the departments. Several AI avatar roles were assigned, from IT security and administration to management, to recreate the workflows as realistically as possible. Our experts provided advisory support throughout.
- 2
The process
Participants interacted directly with the avatars, made decisions in real time and went through the scenarios in a playful, interactive way. Every simulation was moderated, evaluated and could be repeated at will.
- 3
The result
The teams gained confidence in IAM and encryption processes, weaknesses in workflows and communication became visible, and the security level was noticeably raised.
AI-VI Core Technology
AI-VI is built on a proprietary Core Technology that doesn't simply process content. It structures, validates and delivers it as interactive video dialogues.
Content Upload
Presentations, documents and videos are ingested and automatically analysed.
Structuring & Dialogue
Structured questions and answers are generated and prepared for approval.
Review & Approval
Content is reviewed, adjusted and explicitly approved. Only then is it available.
AI-Avatar Output
Queries are recognised, matched and delivered as approved avatar videos.
No generative responses. Every output is based on pre-validated content, fully controllable.
What sets AI-VI apart
Controlled rather than generative AI
Responses are based exclusively on validated content. Hallucinations are structurally prevented.
Traceability & Transparency
Every answer can be traced back to specific, approved content.
Human interaction through AI avatars
Information is not just read. It is experienced through realistic AI avatars.
Define once, scale indefinitely
Content is created centrally and made available across the organisation.
One system for multiple use cases
The same technology works for Audit, Knowledge and Simulation.
Our Products
One platform. Multiple possibilities.
AI-VI Audit Assistant
Structured audit preparation through AI-powered video interaction
AI-VI Audit Assistant digitalises and standardises audit processes through structured knowledge preparation and interactive AI avatar videos. Auditors get access to verified content at any time, reducing coordination effort and making processes significantly more efficient.
Explore AI-VI Audit AssistantAI-VI Cybersecurity Coach
Realistic crisis and cyber simulations
The AI-VI Cybersecurity Coach enables interactive simulations of realistic incidents. Employees train decision-making processes in dynamic scenarios and are specifically prepared for critical situations.
Explore AI-VI Cybersecurity CoachAll products are built on the AI-VI Core technology
Why AI-THINK.
We are a German AI software company with our own patent-pending core technology. Our in-house development team in Germany uses modern tech stacks, guarantees fast delivery of your individual requirements and makes your data sovereignty the standard.
Justin Shabani
Managing Director & Founder
Responsible for strategic direction, brand management and communications. Makes AI tangible, understandable and human.
Michael Kraus
Finance & HR
Ensures stable structures, clear processes and financial reliability, and leads acquisition in audits and compliance.
Tobias Nicolay
Senior Developer
Builds scalable architectures and robust APIs. Owns frontend, backend and infrastructure.
Regine Meyer
Financial Accountant
Runs the books and keeps taxes, deadlines and closings in view. Makes sure every figure is documented.
Jan Bechtoldt
Back Office & Executive Support
Handles back-office work and supporting tasks for the management, so the focus stays on what matters.
AI-VI
VirtualInteractive AI avatar · part of the team
Strictly speaking not a team member, and yet somehow exactly that: AI-VI is our core technology in person. She guides audits, moderates crisis simulations and answers questions, in over 70 languages, around the clock, and always only from what you have approved.
There's more behind us. We're backed by a strong network of specialists: compliance experts, software developers and industry insiders. The right expertise for every challenge, on demand.
Questions about the Cybersecurity Coach. Answered briefly.
What security and HR leads ask us most often in a first call.
A training environment where your people play out the emergency before it happens. AI avatars take the roles of everyone involved, the scenario follows your actual processes, and participants decide in real time. None of this is a video to watch, it is a conversation in which decisions have consequences.
We design a realistic scenario with you, for instance an incident in IAM or encryption. The avatars take the roles, our experts moderate live and add pressure where useful. Participants work with the tools they use anyway, from email to video call. Afterwards everything is reviewed and the scenario can be repeated as often as you like.
IT security, management, legal, HR, press and police, depending on who would be at the table in a real incident. That is what separates it from a training deck: the uncomfortable questions come from the direction they would really come from.
Someone who has seen a presentation knows something. Someone who has played through a crisis can do something. The difference shows in the first minutes of a real incident, when responsibilities and reporting lines have to be second nature. Add repeatability: the same scenario, as often as needed, with different teams and comparable results.
It depends on the scope: how many scenarios, how many participants, how often repeated, and whether it runs in our environment or in yours. So there is no off-the-shelf price list, there is an offer that fits your training needs. The fastest way to clarify it is a joint call, that is enough for a reliable estimate.
An exercise takes about 90 minutes instead of two days, because neither rooms nor external actors are needed. In one real-world case that made more than 200 employees trainable, saving roughly €300,000 a year compared with the classic format. Most of the preparation sits with us.
Platform, documents and the AI processing of text sit on servers in Germany. To generate the avatar videos we use a specialised provider in the USA by default. Only the approved answer texts go there, plus the recordings of the person shown in the case of a custom avatar. This is covered by EU Standard Contractual Clauses. Your content is not used to train AI models. In the Enterprise plan it can run inside your own infrastructure. By default, the avatar videos are still generated by the provider in the USA in that case. You receive the evaluations, while scoring individual people is explicitly not the purpose.
What runs where? What Provider Location Application, databases, file storage, mail server netcup GmbH Data centre in Nuremberg, Germany Encrypted backup netcup GmbH Data centre in Vienna, Austria AI processing of text, our own models Trooper.AI (Media Trooper GmbH, Idstein) Data centres in Germany Generation of the avatar videos Specialised provider USA (Amazon Web Services, region us-west-2) covered by EU Standard Contractual Clauses - Only the reviewed and approved answer texts go to the video provider. For a custom avatar, the image and audio recordings of the person shown are added, and that person has given written consent beforehand.
- Our own models process texts on servers in Germany. Uploaded documents do not leave the EU.
- The hosting provider netcup GmbH is certified to ISO/IEC 27001:2022 (TÜV NORD, valid until 17 November 2027), not AI-THINK. itself.
With one conversation. In the demo we run a simulation live and you see straight away how it feels for your teams. Many then start with a scenario from their own risk profile, reviewed and repeatable, with no long-term commitment.
Anything left open? Write to us
Test your crisis resilience before reality does.
Controlled, auditable AI. Made in Germany. 🇩🇪