Knowledge base
The curated, indexed set of corporate documents your assistants answer from.
TL;DR
Curated company documents indexed for assistants.
Quick facts
- Category
- AI & knowledge management
- Product tie-in
- KobiGPT RAG platform
- Related
- See compare and tools pages
- Locale
- TR and EN site
Why teams choose KobiGPT
- Understand terms before evaluating vendors.
- Link concepts to KobiGPT features (RAG, Kobi Kredi).
- Share glossary links with procurement and legal.
- Explore assistant use cases next.
Product facts
- Ücretsiz plan
- 100 doküman · 2 departman · 120 Kobi/ay(PLAN_CONFIG)
- Starter
- 1000 doküman · 5 departman · 1000 Kobi/ay(PLAN_CONFIG)
- Pro
- 12500 doküman · 25 departman · 12500 Kobi/ay(PLAN_CONFIG)
A knowledge base is not an archive
A knowledge base is not where every company file is dumped; it is a selected set. Indexing every file does not raise answer quality, it lowers it — irrelevant content adds noise at the retrieval step and can crowd out the right passage.
The selection criterion: does this document contain the answer to a repeatedly asked question, and is its content current? If both are yes, it belongs in scope. One-off files, superseded material, and anything containing personal data should stay out.
Scope separation and assistant structure
In KobiGPT the knowledge base is not a single pool; each assistant has its own scope. Files attached to the HR assistant are not present in the accounting assistant's search space. That both improves precision and forms the basis of access control.
Plan limits bound the structure: the Ücretsiz plan plan covers 2 assistants and 100 documents, and the Pro plan 25 assistants and 12500 documents.
Maintenance: the hardest part
Building a knowledge base is easy; keeping it current is hard. When a procedure changes, uploading the new file is not enough — the old one must be taken out of scope. With both versions attached, the assistant may treat each as valid and produce contradictory answers.
The practical recommendation is to review the attached file set at regular intervals. Citation doubles as an audit tool here: if an unexpected old file appears as a source, that is the signal it should no longer be in scope.
Applying A knowledge base is not an archive in a controlled workflow
A useful way to evaluate knowledge-base is to follow one real question from the source document to the final answer. Record which file was selected, what context reached the model, and what a reviewer would need to verify. This turns a definition into an operational check and makes the result comparable across teams.
The same check should include ownership and change management. Decide who updates the relevant documents, how an outdated result is reported, and which access boundary applies. KobiGPT can provide the assistant and the cited document context, but the organisation still owns the source material, permissions, and the decision made from the answer.
FAQ
What is Knowledge in practice?
Curated company documents indexed for assistants.
Does KobiGPT use this?
See product docs and feature pages for implementation details.
More reading?
Visit our blog and FAQ.
Accuracy disclaimer?
Educational content; verify for compliance decisions.
Is it better to upload every file?
No. Irrelevant content adds noise at retrieval and lowers precision. A curated set performs better.
How do we manage old versions?
Take superseded files out of scope; otherwise the assistant may cite the old and new versions at the same time.
Comparison
| Feature | KobiGPT | Alternative |
|---|---|---|
| SME focus | Yes | N/A |
| Citations | When using RAG | N/A |
| Glossary depth | Growing | N/A |
| Tools | Interactive | N/A |