Knowledge, not chat history
Approved sources, traceable answers and roles instead of uncontrolled individual tools.
PaulinAIAn HDN product↗PaulinAI · HDN's own AI platform
Company knowledge becomes actionable.
PaulinAI combines approved knowledge, specialised AI agents, interfaces and human approvals in a controlled workspace.
HDN AI · PROCESS INTELLIGENCE IN MOTION
Agents with boundaries
Agents perform defined tasks while sensitive decisions stay with people and are logged.
Integration, not isolation
APIs and controlled handovers connect PaulinAI to existing systems and processes.
Deployment models
Cloud, private or local – aligned with your protection needs.
The architecture is assessed against data classification, integrations, performance and internal requirements.
Managed cloud
Fast entry with a centrally managed platform and clear tenant and permission concepts.
For standard requirementsPrivate environment
Dedicated components and controlled data flows for increased requirements.
After technical reviewLocal deployment
Selected components can run in your environment when infrastructure and models fit.
Individual architectureWhen PaulinAI fits
A platform – or custom development?
- ✓Internal knowledge must become securely accessible.
- ✓Teams need shared agents, roles and approvals.
- ✓Data sources and workflows should be connected in a controlled way.
- ✓For highly specific processes HDN can build a custom app or agent solution instead.
Frequently asked questions
Clarity before the first step.
No standard demo and no unnecessary technology overhead. We start with your process and the required safeguards.
What is the PaulinAI Hub?+
The PaulinAI Hub is the central, protected workspace for company knowledge, P-Apps and automation. P-Chat, data sources, workflows, permissions and controlled browser or desktop actions come together in one platform.
Which organisations is Paulina designed for?+
Operating dedicated local AI is particularly valuable for medium-sized companies, industrial organisations and larger enterprises with sensitive knowledge, multiple departments or recurring processes. Paulina is modular, so an organisation can begin with one defined use case and expand the platform step by step.
Can PaulinAI Hub run entirely on-premises?+
Yes. Knowledge processing, search, storage and local models can run in your own infrastructure. External AI services remain optional and explicitly controlled.
How do AI permissions work?+
Roles, groups and access rights can be defined down to each folder, database, mailbox or website. The permission filter runs before retrieval and before content enters the AI context, so users only receive information from sources approved for their groups.
What happens when a sensitive source is used?+
The sensitive-data guard disables cloud AI, external model providers and web search immediately for that chat. Processing remains local and the classification persists for the entire conversation.
Can Paulina operate software without an API?+
Yes. Paulina can operate web and desktop interfaces directly. Every relevant action is logged and actions with external impact require approval.
What is P-Chat?+
P-Chat is Paulina's central AI command centre. It brings authorised company knowledge, analyses, diagrams, workflows, P-Apps and controlled browser or desktop actions into one conversation.
Can departments build workflows themselves?+
Yes. P-Workflow provides more than 100 admin-approved building blocks for triggers, data access, processing, AI, loops, approvals and actions. Critical steps can be tested first and require confirmation before they take effect.
Which data sources can Paulina connect to?+
Examples include file systems, MySQL, MariaDB, PostgreSQL, MS SQL, ODBC and Access, protected Oracle access, IMAP, Microsoft Graph, REST services, websites, AD or LDAP and Microsoft Office. Access is controlled for each source and user group.
How is Paulina kept up to date?+
The PaulinAI environment typically runs on a virtual Linux machine controlled by the customer. Approved releases can be installed through the update server, while P-Apps, connectors and customer-specific automation can be added gradually.
How does a PaulinAI project start?+
We begin with a focused process review. Together we define the expected value, connected systems, data, security boundaries and a sensible first production workflow.
