FAQ
Frequently asked questions.
Short answers.
About the platform, the agents, your data and how we work together. If your question is missing, ask it in the demo.
01Platform
What the platform is.
01What is the LumeSec AI Platform?
The operating layer on which you describe your business objects and processes. The platform runs them end to end: it reads documents, writes documents in your style and executes steps. Your team approves. Identity, permissions, models and audit trail are handled once by the platform. Every agent on top inherits them.
02What do we have to describe?
Three things: your business, your objects and your processes. That is what you do, what you work with (for example customers, files, invoices, policies) and which steps cost time today. From this the platform builds the objects, the processes and the documents. You review and approve.
03Does the AI calculate our figures?
No. The AI reads and explains. Figures are calculated with fixed formulas in a rule engine. The same inputs give the same result, and you see the calculation for every value. Your team decides.
04Does the platform replace our AI models?
No. The platform connects models, for example from OpenAI, Anthropic, Mistral or Google, open models such as Llama, or models in your own house. It controls who may use which model with which data.
05Can we use the platform without your interface?
Yes. Everything the platform does is also available through an API. You can bring the results into your own applications.
02Agents
What the agents do.
01Which agents are there?
The Financial Services Agent for every financial services topic in banks, from statement analysis to account and credit processes, and the Kanzlei Agent for tax advisory firms. Next to them are agents for credit applications, legal and InfoSec. For your own process we build an agent on the same platform.
02Who is the Kanzlei Agent for?
For tax advisory firms in Austria. It reads documents and exports from RZL and BMD, carries over values with their source, calculates ratios and builds the closing presentation in your firm's layout.
03How do we teach an agent our rules?
Your team describes the rule in plain words. A second agent, the Architect, turns it into a proposal. The rule is only adopted when your team agrees. Every rule is versioned. Nobody has to write code.
04Who makes the decision?
Always a person. The agent delivers proposals and drafts. Nothing changes and nothing leaves the house without approval from someone on your team.
03Data & security
What happens to your data.
01How do we know where a value comes from?
Every value names its source: document, page and quote. Calculated values show the formula and the inputs.
02Do you train models on our data?
No. We do not train models on customer data.
03Which permissions does an agent have?
The rights assigned to it. You set what an agent may read and do. It has no direct database access.
04What is logged?
Who did what and when, and with which source. The audit trail is stored so that it cannot be changed. It serves as evidence for internal audit and supervisors.
05Which frameworks does LumeSec cover?
We map requirements from the EU AI Act, DORA, NIS2, ISO 27001 and the GDPR onto the platform and support you with the evidence. The legal assessment stays with you and your advisors.
06How do we report a vulnerability?
Write to security@lumesec.ai. You find the PGP key on the contact page. We acknowledge every report within one working day and disclose in coordination, by default after 90 days.
04Deployment
Where the platform runs.
01Where does the platform run?
In one of three variants. SaaS: hosted by us in Europe. Dedicated: an environment of your own with its own data storage and its own keys. Sovereign: in your data centre or your cloud, without outside connection if required.
02Is our data separated from other customers?
Yes. Every organisation has its own environment with its own roles and permissions. Sign-in runs through single sign-on, with a second factor or passkey.
03Which systems do you connect?
For example Entra ID, Okta and Keycloak for sign-in, Microsoft 365, SharePoint and Teams, RZL and BMD in tax advisory. We connect your line-of-business systems through interfaces. Which ones, we clarify in the demo.
05Working together
How we work together.
01How does the demo work?
30 minutes with Marcus Ziegler, the founder. You bring a process, we show it live on the platform. The demo is free and without obligation, in German or English.
02What happens after the demo?
We set up the first process together with your experts and test it on your real cases. Then it goes into production. The next process follows once the first one runs.
03What do we contribute to the setup?
One person who knows the process. Three to five real cases for testing. Your policies, checklists and templates. We do the setup.
04What does it cost?
After the demo you get a clear figure for your case. We do not publish prices because process and deployment determine the effort.
05Do we need our own AI experts?
No. Your team describes rules in plain words. We take care of the technical side.
06Who owns the rules?
You do. Your rules, mappings and templates belong to your organisation.


Marcus M. Ziegler
Founder & CEOBirgitz, Tyrol--:--
A demo directly with Marcus.
30 minutes on your example. Marcus shows you what the LumeSec AI Platform can do. Afterwards you know what an agent takes over for you.
30 min · Microsoft Teams · German or English
We show: the analysis of one set of annual accounts up to a draft vote