Skip to main content
+49 228 84239000
+49 228 84239000
Nexoria
Get started

Nex · Your AI assistant

Nex: the AI assistant for your company knowledge, within your access permissions

Ask questions of your company knowledge. Keep working with sources and trigger tasks directly from the conversation.

Discover Nex

ChatGPT alternative · AI chat

Nex AI Assistant in everyday work

One question. Your company knowledge as the answer.

Nex turns shared content into answers, summaries and next steps.

ChatGPT alternative · AI chat

Check answers against sources

Nex compares maintenance contracts; you open the cited evidence and check the statement against the original.

Query knowledge.

Nex answers based on the shared content of your workspace.

Trace every source.

References to the documents used make answers verifiable.

Keep working.

The conversation turns into tasks, appointments and new documents.

Nex · Company knowledge

Your question.
Your company knowledge.

  • Ask questions of the indexed content in natural language.
  • Summarise and analyse long documents and attached files.
  • Trace the documents used via the source references.
View all features

Nex · Everyday work

From the conversation
to the next step.

  • Create tasks and appointments directly from the chat.
  • Create presentations, reports and document comparisons directly in the Nex AI chat.
  • Enable internet research when needed and use it with sources.
How Nex works
Creating a presentation in the Nex AI chat with anonymised sample content.
Creating a presentation in the Nex AI chat with anonymised sample content.Enlarge view

Nex · Permissions

Answers within the scope
of your permissions.

  • The workspace limits the accessible knowledge.
  • Permissions are checked before content is accessed.
  • The permission logic also applies to search and chatbots.
Understand the permission logic

How does Nex work?

If you wish, internet research can also be enabled. The search infrastructure for this is operated in Germany, and these results are also shown with their source.

  1. Step 1

    Question in natural language

    Questions are asked in the chat, without search syntax and without needing to know in advance which document contains the answer. Documents can also be attached to the question.

  2. Step 2

    Search in the shared content

    Nex searches semantically in the indexed content of the workspace in which the question is asked. It is based on the same index that document search uses.

  3. Step 3

    Answer with source citation

    The answer is built from the passages found. The documents used are identified, so that the statement remains verifiable within the company rather than standing as an unsupported claim.

  4. Step 4

    Further processing, not just answers

    From the answer, you can generate summaries, create tasks and appointments, create presentations and reports directly in the AI chat and trigger automations.

Nex · Your AI assistant · Process
01of 4 steps
  1. 1Question in natural languageThis step is open
  2. 2Search in the shared content
  3. 3Answer with source citation
  4. 4Further processing, not just answers
Explanatory graphic · Process for the selected step

Nex · Your AI assistant in detail

Everything you want to know.

Nex is Nexoria's AI assistant and also the platform's main interface. It answers questions from the indexed company knowledge of the respective workspace and names the documents used. The permission logic is decisive: Nex works exclusively with content that the user asking is allowed to see anyway. Answer generation runs via Mistral AI, headquartered in Paris.

Definition

What is Nex?

Nex is the AI assistant and main interface of Nexoria; it works with the shared company knowledge of the respective workspace and creates documents, presentations and reports directly in the chat.

Nex is therefore not another tool alongside the platform, but the point where the other building blocks come together. The content comes from Nexoria Drive, and found via the Document search, processed further directly in the AI chat, for example into documents and presentations, and through the automations. The main interface is the Nex AI chat. Document search is also available as a separate search window outside the chat.

Problem

The problem

General AI chats are strong on language and blind to how the business operates. They know neither the company's price list nor the latest version of the framework agreement. They only become useful once someone finds the right documents and attaches them. The person then does exactly the step that was supposed to be taken off their hands, and pushes company documents file by file into an external environment.

The obvious alternative, simply pointing an AI at all company data, creates the opposite problem. An assistant that answers everything for everyone hands out payroll lists, draft termination letters and client files to people who have no access to them. No permission model in a company is designed to be bypassed by a chat window.

At this point, Nex is deliberately built more narrowly than would be technically possible.

Key point

The permission logic: what Nex accesses

This section describes the strongest argument for Nex: Nex respects the access permissions that already apply in your company.

The workspace is the boundary

In Nexoria, access permissions are not set on Nex but on the workspace. A workspace is a knowledge space with members, roles and its own collection of documents and databases. The roles are Admin, Member and Viewer, and checks are carried out on the server side at every endpoint.

Nex inherits these permissions unchanged. It answers a question from the content of the workspace in which it is asked, and from no other. Anyone who does not belong to a knowledge space receives no answer from it, even if the document would answer the question perfectly. The same rule applies to chatbots: a bot receives only the knowledge of the workspace assigned to it.

Why this is a selling point

  • The company's existing permission structure remains valid. No second, parallel permission model is created just for Nex.
  • Sensitive areas such as HR, legal or individual client matters can be run as separate knowledge spaces without switching off Nex for everyone else.
  • For law firms, tax advisers and consultancies with multiple clients, this separation is not optional but a prerequisite for being allowed to use AI at all.
  • Whatever a workspace does not share does not make it into an answer, nor into a presentation generated from it.

What matters during rollout

How the workspaces are structured determines which knowledge Nex makes accessible to which people. Before rollout, therefore, decide which areas stay separate and which knowledge should be widely available. In a conversation, we go through this together with you.

Nex names the sources an answer is based on. This lets your team check directly at the source, before using it, whether a statement is correct and whether the underlying version still applies.

Product data

Nex fact sheet

Product fact sheet

Nex

Nex is the AI assistant and main interface of Nexoria; it works with the shared company knowledge of the respective workspace and creates documents, presentations and reports directly in the chat.

Category
AI assistant for company data
Problem
Replaces general AI chats such as ChatGPT Enterprise or Claude with an assistant that works with your own company knowledge. Slides and reports are created in the chat instead of by hand in PowerPoint or Canva.
Target group
Businesses; the offer is aimed at business customers within the meaning of Section 14 of the German Civil Code (BGB).
Core features
  • Real-time internet research with source citations
  • Access to the knowledge of the respective workspace
  • Summarise and analyse documents
  • Create documents and presentations directly in the Nex AI chat
  • Technical reports and document comparisons with source references
  • Create tasks and trigger workflows
Input
Questions and task descriptions in natural language, attached documents, shared workspace content and, on request, enabled web search
Output
Answers with source references, summaries, presentations as .pptx, reports as PDF, document comparisons, tasks and appointments, and triggered automations
Integration in Nexoria
The platform's main interface, with access to Drive, document search and automations. Creating documents, presentations and reports is integrated directly into the chat and uses only the content shared by the workspace.
Privacy
Answer generation by Mistral AI. In Nexoria’s verified Mistral organisation, the use of new API interactions for model training is disabled (account checked on 26 September 2026).
Hosting
Answer generation by Nex uses Mistral AI.
Price
Included in the core product, token allowances on the pricing page
Status
Feature available. Registration and online purchase of the Base and Pro plans have been available since 15 September 2026.

Requirements

Requirements

  • Nex works with the indexed content from Nexoria Drive. It also answers questions from attached documents and from enabled web search.
  • The workspaces must be structured so that the right people reach the right knowledge spaces. This decision is made by the company.
  • Usage consumes AI tokens. The Base plan includes 3 million tokens per user, the Pro plan 5 million per user.
  • One user corresponds to one device. The same user cannot be used on several devices at the same time.
  • The offer is aimed at business customers within the meaning of Section 14 of the German Civil Code (BGB); all prices are net prices.

Evidence

Once the token allowance is used up, tokens can be topped up automatically or manually. Without a top-up, new AI requests are paused. Document search continues to work and does not consume any tokens.

Source: Pricing. Checked on .

Processing

Data protection and infrastructure

Processing of requests to Nex. As of .
SubjectDetailSource
Model providerMistral AI is a French Société par actions simplifiée headquartered in Paris.Mistral AI, Legal notice
checked on
Location of answer generationAnswer generation by Nex uses Mistral AI.Security and privacy
checked on
Embeddings and searchDocument embeddings and semantic search run on servers in Germany, using models that Nexoria operates there itself.Security and privacy
checked on
Internet researchInternet searches run on our own search infrastructure on German servers. Search queries are not forwarded to Bing or Google.Security and privacy
checked on
Use for trainingIn Nexoria's verified Mistral organisation, the use of new API interactions for model training is disabled (account check of 26 September 2026). This setting applies to new interactions processed there. For your own optionally connected AI accounts, the settings of the respective account apply.Data processing agreement
checked on
Tenant separationStrict data isolation per tenant at database and vector store level, a role model with Admin, Member and Viewer, server-side permission checks on every API endpoint.processing register
checked on

Features

Specific features

  • Questions to the shared company knowledge, answered with references to the documents used.
  • Summary and analysis of long documents, including attached files.
  • Real-time internet research, enabled on request, also with source citations.
  • Create tasks and appointments directly from the conversation.
  • Trigger automations and pass data to connected specialist software.
  • Create presentations, reports and document comparisons directly in the AI chat, with references to the sources.
  • Set answer length, level of detail and language for each conversation.

Which documents and presentations Nex creates in the AI chat, and how they are structured, is explained on the page Documents and presentations in the Nex AI chat.

Comparison

Comparison with the conventional approach

The comparison is with the category of general AI chat, not a specific product. Details on individual providers belong on the comparison pages, with the source and date of review.

General AI chat and Nex. As of .
AspectGeneral AI chatNexSource
Knowledge basisThe language model's knowledge, supplemented by files attached to each requestThe workspace's indexed content, supplemented by attached filesStatement by Nexoria
TraceabilityThe answer stands on its own, and its origin is not readily apparentThe documents used are identifiedStatement by Nexoria
Access permissionsAnyone who can attach the file gets an answer about itThe workspace decides, and the same permissions apply to search and chatbotsWorkspaces and databases
checked on
What happens nextCopy out the result and continue working on it in another toolTask, appointment, presentation, report or automation directly from the conversationStatement by Nexoria
Processing locationVaries by providerSearch and embeddings in Germany, answer generation by Mistral AISecurity and privacy
checked on

Orientation

Where to find what

Sources

As of

  1. Mistral AI, Legal notice. Mistral AI, 2026. https://legal.mistral.ai/legal-notice Retrieved on .Primary source for the model provider's company name and registered office. The year refers to when the page was accessed; the page itself has no publication date.
  2. Regulation (EU) 2016/679, General Data Protection Regulation. Publications Office of the European Union, 2016. https://eur-lex.europa.eu/eli/reg/2016/679/oj Retrieved on .Source for the data processing agreement under Art. 28 mentioned on this page.

As of

FAQ

Frequently asked questions about Nex

Where does Nex get its answers from?

From the indexed company knowledge of the workspace in which the question is asked, from documents attached to the question and, on request, also from internet research. The answer names the sources used, so that the statement remains verifiable within the company.

Does Nex see all of the company's files?

No. Nex works with what the respective workspace shares, and therefore with what the person asking is allowed to see anyway. Anyone without access to a knowledge space will not get an answer from it either. This keeps your company's permission structure valid in the chat too.

Which language model is behind it?

Answer generation by Nex uses Mistral AI. Separately, the embeddings for semantic search are computed locally on Nexoria's own servers in Germany. Other services and enabled integrations are described in the security overview and the processing register. ChatGPT (OpenAI), Claude (Anthropic), Gemini (Google) and Langdock can optionally be connected to Nexoria. Nexoria also works without these integrations. When connected, shared content is transferred to the respective provider. That provider's terms, processing locations and settings apply.

Is our content used for training?

In Nexoria's verified Mistral organisation, the use of new API interactions for model training is disabled (account check of 26 September 2026). This setting applies to new interactions processed there. For your own optionally connected AI accounts, the settings of the respective account apply.

What distinguishes Nex from a general AI chat?

A general AI chat answers from its model's knowledge and from whatever is attached to each request. Nex accesses the workspace's indexed content, respects the company's access permissions and identifies the documents used. It can also create tasks and appointments and trigger automations.

How much does Nex cost?

Nex is included in the core product and costs nothing extra. It is AI tokens that are consumed: the Base plan includes 3 million tokens per user, the Pro plan 5 million per user. Once the allowance is used up, Nex accepts no further requests, while document search keeps running and does not consume any tokens.

Contact

Next step

How much Nex achieves in your organisation depends on how the knowledge spaces are structured. In a conversation, we go through which areas need to stay separate and which knowledge should be widely available.

Arrange a callView all building blocks