AI assistant for specialist and educational publishers
Your content is the product – your systems are the tooling behind it. nerilio makes both conversational: for subscribers searching the licensed catalogue, for editorial and production teams who have to keep house rules and workflows in their heads, and for everyone working with your publishing systems day to day. Answers with a source reference, without general knowledge mixed in, GDPR-compliant on EU servers.
Where knowledge gets stuck in a publishing house
Search returns hits, not answers
Someone with a specific question gets a list of 40 documents. The answer is on page 12 of one of them – and never gets found.
The same thing is explained again and again
What was the approval chain? Where is the rule on image rights? How does this article reach the export? The answers exist already – yet they are still given in person every time.
The comparison product is free
Subscribers now ask a free tool first. It answers instantly, often wrongly and without a source – but it answers. The question of what the subscription is for follows on its own.
Three groups, one knowledge base
The same assistant, with its own catalogue and its own access per audience. You decide who sees what.
Readers & subscribers
The licensed catalogue becomes conversational: asking instead of full-text search, answers with a source reference down to the title. Run behind the paywall, it becomes a tier of its own.
Editorial & production
House rules, style guides, rights clearance, workflow and approval documentation: what is otherwise scattered across handbooks, wikis and emails, the assistant answers in the context of the work. New colleagues ask instead of interrupting someone.
Users of your systems
An assistant on the product documentation of your publishing software – answering usage questions before they turn into a support ticket. Exactly the systems we at SNAP deliver and support anyway.
Your catalogue becomes a conversation
The same content, a different way in. Without reworking, rewriting or migrating any of it.
An answer, not a result list
“What does the handbook say about tax relief for solar installations?” – nerilio reads the catalogue and formulates the answer. The research disappears, the basis for it does not.
Traceable to the passage
On request every answer states which title and which passage it came from. That is precisely what separates your offering from a general language model.
Your catalogue only
nerilio uses RAG technology and answers exclusively from the content you release. No general knowledge, no outside sources, no invented citations.
Internal as well as external
The same approach works for subscribers, for the editorial team and for user support. Separate assistants per audience, each with its own knowledge base and its own access.
We have known the publishing market for 30 years
nerilio is a product of SNAP Innovation in Hamburg. SNAP was founded in 1996 and has been setting up the systems in which publishing content is produced ever since – from editorial through production to delivery. Whoever configures your assistant knows the systems behind it from daily work, not from an integration list.
- vjoon K4 – multichannel publishing system
- vjoon seven – central content hub
- PublishOne – authoring platform
- Sharedien – smart DAM
- dataplan JournalSuite & JournalDesigner – planning
- Twixl Publisher – digital publishing
- axaio MadeToPrint & MadeToTag
- callas pdfToolbox
- enfocus PitStop Pro, PitStop Server & Switch
- EasyCatalog
nerilio is the newest part of that toolbox – and the first one that gives both your content and your systems a voice.
Licensed content stays licensed
For a publisher the decisive question is not what the AI can do, but what happens to the catalogue. So: nerilio processes exclusively on EU servers, your content feeds into no model training, and the assistant only reaches the material you release.
- No training of AI models with your content – contractually secured via Microsoft Azure
- Hosting and processing exclusively in the EU
- Separate assistants for internal and external audiences – editorial knowledge stays internal
- Data processing agreement under Art. 28 GDPR provided as standard
In your portal, under your brand
In your existing portal
A single script tag is enough, the chat widget renders itself. No rebuild of your platform, no release of your own.
On the intranet and in editorial
The same route for internal assistants – as a widget on the intranet or as a page of its own beside the editorial system.
Through the REST API
From the Pro plan onwards you embed nerilio deeply into your application: ask questions, populate the knowledge base programmatically, receive answers with source references.
Fully white-labelled
Your own name, logo and colours, and your own domain on request. To your readers it is your product.
Frequently asked questions from publishers
No. nerilio runs on the Azure OpenAI Service, for which Microsoft contractually guarantees that processed data does not feed into the training of the underlying models. That covers the titles you upload as well as the questions your users ask. The details are in the data processing agreement, which you can review before signing.
Yes, and that is the normal case: one assistant for subscribers with the licensed catalogue, a second for editorial and production with house rules and workflows, a third on the product documentation of your systems. Each has its own knowledge base, and access can be secured via login, SSO or SAML. Subscription tiers can be mapped the same way.
PDF, Word, PowerPoint, Excel, JSON, XML, video transcripts and web pages by URL. On top of that come connections to CMS platforms such as PublishOne, to learning platforms such as Moodle, ILIAS and totara, and to your own systems via the REST API. An updated title is re-indexed automatically.
nerilio answers exclusively on the basis of the content you provide. If information is missing, the assistant says so instead of constructing a plausible answer. On request the source used is shown with every answer, so each statement can be traced back to the title it came from.
A first assistant with your content is ready in minutes: upload, choose the operating mode, embed the widget. It gets more involved with deep portal integration, tiered catalogues or full white-labelling – and there you are supported by the team that already knows your publishing systems.
Hear your own catalogue speak
In a short demo we show nerilio using your own titles or your system documentation – and you see what an answer with a source reference looks like on your content.