Glossary: the terms behind the AI knowledge base
Every term explained in 60 words or fewer, no technical background needed. Easy to look up, share and quote – every entry has its own anchor link.
AI knowledge base
An AI knowledge base is a central, searchable store for company documents, combined with an AI that answers questions about them in plain language. The answers are based on the company’s own documents and show the source. It is designed for documented know-how – instructions, reports, solutions to problems – not for records that stay in the ERP.
RAG (retrieval-augmented generation)
RAG is a method in which an AI, before answering, first retrieves the relevant passages from a collection of documents and bases its answer on them. This reduces invented answers (hallucinations) and makes it possible to cite sources, so each answer can be checked. RAG is the technical basis of most AI knowledge bases.
LLM (large language model)
An LLM is a large AI language model that understands and writes text – well-known examples are the models behind ChatGPT or Claude. On its own, an LLM knows nothing about your company documents; only when combined with them (see RAG) does it become a tool for internal company knowledge.
OCR (text recognition)
OCR (optical character recognition) converts images of text – scans, photos, old PDFs – into searchable text. For companies, this means documents from the paper archive can be included in the search too. How well it works depends on the original – print quality, scan or photo quality, age; handwriting is the hardest case.
Semantic search
Semantic search finds content by meaning rather than exact wording. A question about “heating not getting warm” can also bring up the report that says “flow temperature too low”. Classic full-text search, by contrast, needs the exact word used in the text. Semantic search is a core component of most AI knowledge bases.
Embedding
An embedding is a text converted into a series of numbers that represents its meaning. Similar content gets similar series of numbers – this lets a system calculate how close in meaning texts are. Embeddings make semantic search technically possible and are stored in a vector database.
Vector database
A vector database stores embeddings (series of numbers representing meaning) and retrieves the passages that are closest in content to a question. It is the “memory” of an AI knowledge base: with every query it supplies the relevant document excerpts from which the AI builds its answer.
Hallucination
A hallucination is a convincingly worded but factually wrong AI answer. It happens when a language model answers without reliable sources. The remedy for businesses: ground the AI in your own documents (RAG) and show the source for every answer, so it can be checked against the original.
Grounding
Grounding means that an AI answer is traced back to specific sources in real documents, for example via a numbered reference for each statement. For companies, this is the difference between “sounds plausible” and “can be verified”: the source lets you open the original document and read it yourself.
Tagging
Tagging assigns attributes to each document – customer, project, installation, date. The AI suggests them when a document is added; a person confirms the document type, because it decides who can see the document. Tagging makes large collections filterable and helps match answers to the right customer or project.
DMS (document management system)
A DMS stores documents in a structured way and makes them easy to find again using folders, metadata and keywords. To find anything, you need to know how things are filed. An AI knowledge base goes one step further: you ask a question in plain language and get an answer with its source.
DPA (data processing agreement)
A data processing agreement under Art. 28 GDPR governs how a service provider processes personal data on behalf of a company – for example when hosting a knowledge base. It is required by law and sets out the purpose, security measures and sub-processors. It’s worth asking for a copy and reading it before you sign.
EU AI Act
The EU AI Act is the European law on artificial intelligence. It divides AI applications into risk classes and applies in stages: transparency obligations (Art. 50) since 2 August 2026, high-risk obligations from December 2027 or August 2028. Companies that use AI must take steps to ensure their staff are AI-literate (Art. 4).
Tacit knowledge
Experiential knowledge is the know-how in the heads of long-serving employees: how a system is set up, how a problem was solved at a customer’s site, what to watch out for on a project. It is lost when people retire or leave – unless it is documented and made searchable.
Fine-tuning
Fine-tuning means training a language model further on your own texts. For company knowledge, this is rarely the right approach: the knowledge becomes part of the model, individual items can’t be deleted, there are no permissions and no sources are shown. A knowledge base using RAG keeps the documents separate – they can be changed or deleted, and answers can be traced back.
Prompt
A prompt is the input to an AI system – in a knowledge base, simply the question someone types or dictates. Good questions name the specifics: customer, installation, project number or time period. No special jargon is required; you ask the way people talk at your company.
Token
A token is the smallest unit of text a language model works with – roughly part of a word. The number of tokens determines how much text a model can process at once and how much processing costs. That is why a system retrieves the relevant passages instead of reading entire folders.
Context window
The context window is the amount of text a language model can take into account for a single answer. It is limited, so no model can read an entire document collection. The system first retrieves the relevant passages and puts only these into the context window – that is the core of the RAG method.
Chunking (text sections)
Chunking means splitting long documents into smaller sections before they are made searchable. Where the sections break affects answer quality: too small and they lose context, too large and the source becomes vague. That is why we decide during set-up how best to divide contracts, reports or drawing labels.
Metadata
Metadata is the set of attributes of a document: customer, project, installation, date, document type. It makes large collections filterable and ensures that an answer relates to the right case. When a document is added, the system suggests most values from its content; a person confirms the document type.
Document type
The document type says what something is: quotation, inspection report, payslip, contract. It is the most important classification, because visibility depends on it – anyone allowed to see a document type sees all documents of that type. That is why the system only suggests the type and a person confirms it.
Roles and permissions
The permission concept sets out who can see and search which documents. It works through roles and document types, not individual files – otherwise it would be impossible to maintain day to day. A cut-off date is also possible: only documents from or up to a certain date. If a role lacks permission, the system says so.
Full-text search
Full-text search finds documents containing the exact words you type. It is fast and precise, but depends on wording: a search for “maintenance contract” does not return a “service contract”. A knowledge base therefore combines full-text search with semantic search, which looks for meaning rather than exact words.
Indexing
Indexing is the process by which documents are made searchable: recognising text, splitting it into sections, assigning attributes and adding it to the search index. It runs in the background as soon as a document is added. Only once a document has been indexed can it appear as a source in an answer.
Interface (API)
An interface is a channel provided by the software vendor that lets other programs read its data. Whether a program has one, and whether it is enabled and reachable on the network, depends on version, licence and set-up. We check this before every connection – sometimes a scheduled export is the more robust option.
Database export
A database export is a scheduled, recurring extract from a business application: at fixed intervals, the program writes files to a folder from which they are read. This is used when there is no interface. It has one advantage over a live connection: nothing changes in the live system.
On-premise
On-premise means a program runs on a server at your own company rather than on the provider’s servers – as many long-established industry-specific programs do. junova itself always runs on our servers. Files from such a server are encrypted and mirrored via outbound connections only, by a small read-only component on that server; no access from outside is opened up.
Separate instance
A separate instance is a distinct installation per company – its own database, its own search index, its own logins. There is no shared system in which several customers’ documents sit side by side. It also makes things simpler at the end: when the contract ends, the documents are exported and the instance is deleted.
ERP
An ERP is a company’s commercial software: orders, purchases, stock, invoices. It knows WHAT was sold and invoiced. A knowledge base is for HOW something was done – instructions, reports, solutions to problems, correspondence. It does not replace an ERP; where the ERP can be connected, it is only read and nothing is changed in it.
Controller and processor
Under data protection law, the company is the controller: the data belongs to it, and it decides how it is used. The provider of the system is the processor and may only do what is agreed in the contract. This is governed by the data processing agreement (DPA), which sets out the instructions, the security measures and deletion.
Deletion policy
A deletion concept sets out what is deleted and when. In practice, deleted items are removed from the search index, so they no longer appear in answers, and are permanently deleted after a set period – or immediately on request. Without the first step, a deleted document could still surface in answers even though nobody can open it.
Knowledge management
Knowledge management is how a company handles what its people know: documenting knowledge, keeping it findable and passing it on. The hard part is rarely writing it down; it’s finding it again years later. That is where a knowledge base comes in: rather than requiring a new filing system, it answers questions from the documents you already have.
Shadow IT
Shadow IT means tools that employees use without approval because the official ones fall short – for example company documents that end up in a private AI account. The problem isn’t the people but the gap: if there is no approved option, people find their own. An approved system with permissions can close this gap.
Employee AI literacy
Since Regulation (EU) 2026/1744, the EU AI Act requires companies using AI, small ones too, to take steps to ensure staff AI literacy (Art. 4). No specific level or certificate is required – a documented basic understanding is: what the system can and cannot do, and what does not belong in it. A usage policy and training records are the usual evidence.
Clear on the terms, but not the benefits?
In a 30-minute video demo we use sample documents from your industry to show what the answers look like, with sources.