AI platform or enterprise search tool?

What is an enterprise search tool?

Enterprise search (also called an insight engine) makes an organisation’s knowledge searchable. It reads in documents from many systems, such as file shares, document management, wikis, email and specialist applications, and stores them in a shared index. Anyone searching gets hits from all sources they have access to. Many providers now answer questions with an AI model and show where the information was found.

Examples from the German-speaking region, according to the providers (as of 11 October 2026):

Provider	Focus	Operation	Pricing basis
Mindbreeze InSpire (Linz)	Insight engine, more than 500 data sources	SaaS, appliance in your own data centre, cloud	Documents in the index, from €83,000 a year
IntraFind iFinder (Munich)	Search across more than 80 connectors, AI answers with iAssistant	On-premises, cloud, SaaS	on request
amber (Aachen)	AI search, chat and agents across more than 120 connected systems	T Cloud in Germany and the EU, in the Enterprise package also on-premises	from €12.99 per person per month
INTERGATOR (Dresden)	Enterprise search with an AI edition	On-premises, private cloud	on request

Other providers of this kind are Sinequa and Squirro.

What is an AI platform?

An AI platform starts from the tasks of staff: drafting letters, summarising documents, translating, answering questions about regulations and guidelines. It combines a chat with ready-made apps for recurring tasks, with permissions from your identity management and a log. Knowledge from the organisation is added through connectors to its systems. More in the guide “What is an AI platform?”

How do the two differ?

Feature	Enterprise search	AI platform
Starting point	Finding knowledge	Getting tasks done
Typical result	List of hits or an answer with references	Draft, summary, translation, answer with source
Data sources	Very many systems and file formats in one index	The most important systems through connectors, plus the person’s own documents
Recurring tasks	Usually not the focus	Apps that are shared and set up for one task
AI model	Depends on the provider: own model at the customer’s site, the provider’s model or external services	Depends on the provider; at basebox, basebox runs the model itself
Pricing basis	Often by data volume or on request	Often per person or per server

The boundaries are blurring: enterprise search providers are adding chat and agents, and AI platforms are connecting more and more systems.

Which questions help with the decision?

• How many sources need to be searched? Ten file stores or hundreds of systems with millions of documents?
• What should staff have in hand at the end? A reference or a finished draft?
• Where does the AI model run? With the tool provider, with an external model provider or in your organisation?
• How do permissions get into the answer? Do the access rights of the source systems also apply to AI answers?
• How is it billed? By data volume, by people or by servers?

More questions are listed in the checklist for AI providers.

How does basebox work with an organisation’s knowledge?

basebox is an AI platform, not enterprise search. Connectors link basebox to your systems, such as SharePoint, Exchange, Nextcloud or Confluence. Every access runs with the permissions of the person asking, read-only by default, and every call to a connector is recorded in the audit log. Every answer shows where it comes from. All connectors from our library are included in the licence; if a system with an interface is missing, we build the connector for a fixed price between €1,900 and €13,800.

basebox runs the AI model itself: in the cloud on our servers in the Munich East data centre, with a separate environment for each organisation, or on a server dedicated to you, at your site. The cloud costs €490 to €178 per person per year (from 5 people, list price, net); the calculator works out any other number.

What does enterprise search offer that basebox does not?

• Very many sources: Mindbreeze names more than 500 data sources, amber more than 120 connected systems, IntraFind more than 80 connectors and over 600 file formats. The basebox library is smaller.
• Search as a list of hits: if you are looking for specific documents and not only an answer, a search interface is often faster.
• Large indexes: enterprise search is built for millions of documents and charges accordingly.

What fits whom?

• Enterprise search fits organisations whose main problem is finding knowledge across many systems.
• An AI platform like basebox fits organisations that want to relieve staff with texts, documents and recurring tasks, with knowledge from the most important systems and the AI model in Germany or on their own premises. More on Connectors.