On-Premises AI Chatbot — Full Control Over Your Data
For companies where data sovereignty isn’t an option but a requirement, Zentor App offers full on-premises operation. The platform runs on your own infrastructure — and with local AI execution, that also applies to the AI models themselves.
In industries such as healthcare, financial services, the legal sector, or public administration, data is subject to special statutory protection rules. Here it isn’t enough to trust a cloud provider — external data transfers are often not even legally permitted, or they require elaborate contractual foundations that are difficult to enforce.
On-premises operation solves this problem at its root: when the software runs on your own infrastructure, the data never leaves your environment. There are no API calls to external services, no telemetry data, no log forwarding to the provider. You are the data controller — and you have the full technical control necessary for that.
In a Zentor App on-premises installation, the entire platform — backend, database, omnichannel inbox, knowledge base, chatbot engine — runs on your own servers. You get access to the deployment artifacts and set up the environment in your own data center or private cloud.
Zentor App provides support with the initial installation and configuration as part of the individual onboarding process. After go-live, operational responsibility lies with you — which precisely matches the goal of full self-responsibility that regulated companies pursue.
Local AI operation refers to running language models on your own hardware. The models run entirely on your own infrastructure — without any connection to external AI services. For a Zentor App on-premises installation, this means: even the AI intelligence of your chatbot never leaves your infrastructure.
Every request your chatbot makes to the language model — whether classifying a customer inquiry, generating a response, or extracting information from the knowledge base — is processed locally. There is no HTTP call to external AI services, no data transfer, no external processing partners.
Zentor App natively supports local model execution as an AI backend. The configuration — which model is used, how it’s parameterized — is entirely in your hands.
All customer data, conversations, and knowledge base entries remain, without exception, on your infrastructure. No data transfer, no external processing partners.
Local AI execution runs on various hardware types. An outage of cloud AI services has no effect on your chatbot’s availability.
On-premises operation is the strongest technical measure for meeting data protection requirements — without relying on contractual assurances.
You integrate Zentor App into your existing security structures: your own firewalls, network segmentation, patch cycles, and monitoring per your standards.
Your platform availability doesn’t depend on the availability of external cloud services. SLA requirements can be met internally.
On-premises AI chatbots are especially relevant for industries where data protection carries particular regulatory weight:
Both options offer a high level of data protection — the difference lies in the degree of control and operational effort:
- Managed infrastructure by Zentor App
- Data on EU servers; external AI connection only via EU Standard Contractual Clauses
- Low operational effort for you
- Available from the Starter plan
- Updates and patches automatic
- Entirely on your own hardware
- Absolute data sovereignty
- Your own security policies applicable
- Local model execution for your own AI models
- Custom plan only
On-premises is part of the Custom plan, which also includes a range of additional features for demanding enterprise environments:
- 10 users included, 1 tenant included (both expandable)
- Local model execution for your own language models and optional Vector RAG
- Advanced authentication options by individual arrangement
- Custom integrations via the Zentor App API
- Individual support agreement and defined response times
- Dedicated onboarding and technical support for the initial installation
For companies with very high message volumes and growing teams: more on scaling and load distribution at the Enterprise AI chatbot for high request volumes.
Frequently Asked Questions About On-Premises Operation
What is on-premises operation?+
On-premises (on-prem for short) refers to running software on a company’s own hardware and infrastructure — in its own data center, private cloud, or on dedicated servers. Unlike cloud solutions, no data flows to external providers: everything remains under your full control.
What is local AI operation?+
Local AI operation refers to running language models on your own hardware. AI processing takes place entirely on your own servers — without any connection to external AI services. Zentor App natively supports local model execution as an AI backend for on-premises installations.
What prerequisites do I need for on-premises?+
For an on-premises installation you need your own server infrastructure with sufficient resources for the Zentor App platform and — when using local AI models — for local AI processing. The exact hardware requirements depend on the number of tenants, expected message volume, and the AI models chosen. These are determined as part of the onboarding for the Individual plan.
From which plan is on-premises available?+
On-premises operation is available exclusively in the Custom plan. The Custom plan also includes 10 users (expandable on request) and 1 tenant included (additional tenants at €39/month each), custom integrations, advanced RAG configuration, and advanced authentication options by individual arrangement — up to €899/month, quote before payment.
Can I use my own language models?+
Yes. In the Custom plan with local model execution, you can freely choose which language models to use — provided they can run on the chosen local AI infrastructure. You are not tied to a specific model. This gives you the flexibility to select models based on performance requirements, licensing, or specific domain suitability, and to switch as needed.
Interested in On-Premises Operation?
Talk to us about your requirements. In a demo call, we’ll clarify whether and how Zentor App On-Premises can be configured for your environment.
Further reading:
More in the Help Center