Skip to content
Waitlist

Knowledge base: bring in files and previous work.

Strategy papers sit in the team folder. Audience analyses are buried in a presentation. Earlier keyword lists live in a spreadsheet. This knowledge already exists – the knowledge base makes it usable for BrandZaps.

So an analysis does not start from zero or repeat work your team has already done.

Material 4 files · 1 note
Drag files here
TXT MD CSV TSV XLSX DOCX PDF
Positioning_2026.docx Text extracted
Keywords_Legacy.xlsx Text extracted
Audiences_Deck.pdf Images skipped
Public research · Humboldt Research running

Knowledge that just sits somewhere.

Company knowledge quickly spreads across documents, spreadsheets and individual heads. A new tool does not automatically know about these decisions.

The knowledge base brings these sources into the project context. It does not turn existing material into new truth, but into an accessible basis for the features that work with it.

Without a knowledge base
Recommendations ignore positioning decisions already made.
An earlier audience analysis gets researched all over again.
A scored keyword list sits disconnected next to the next analysis.
With a knowledge base
Existing material sits in the project context and is available to analyses.
The origin of a statement stays labelled as material.
Later steps build on knowledge brought in earlier.

Three ways to bring knowledge into BrandZaps.

01

Notes for what is not written down anywhere else

Not every important decision exists as a finished document. In the notes field you record previous strategy, positioning and audiences – including connections that are not in any file.

That way an important requirement reaches the project without someone first having to copy it into a new document.

NOTE
Positioning 2026: focus on agencies with 5–30 employees. No enterprise tenders. Audience “in-house team” stays secondary but is still served in content. Free text · can be extended at any time
02

Files via drag and drop

You can drag existing documents straight into the knowledge base. BrandZaps extracts the text and stores it as material. Images inside a file are deliberately not processed.

This limit matters: an embedded image is not silently treated as information that has been read.

SUPPORTED FORMATS
TXT MDPlain text and Markdown
CSV TSVLists and exports
XLSXSpreadsheets with keyword sets
DOCXStrategy and positioning papers
PDFAnalyses and presentations
Images in files are not analysed
03

Public company research with status

Humboldt analyses public mentions and turns them into a cleaned-up summary. The research runs in the background, and its status stays visible.

The finished summary gets a capture date. If a run fails, you repeat it with one click.

Research runningin the background
Research not completedrestart
Summarycaptured 12 Sep 2026

Organised by month instead of an endless material list.

Material is filed in monthly folders. One click opens a month as its own view with file cards. This creates a chronological order without you having to sort every document by hand.

A file's full text only loads when you open it. The overview stays light and shows the cards first rather than the entire content of every document.

09 / 20263
Positioning_2026.docx
Keywords_Legacy.xlsx
Note
08 / 20262
Audiences_Deck.pdf
Market_notes.md
07 / 20261
Humboldt summary
06 / 20261
Competitor_list.csv

Where your material keeps working.

The knowledge base is not a database for its own sake. It feeds knowledge into the features that process it – with visible origin.

In the company profile draft

Material can feed into the first draft and is labelled as material in the evidence. If new material is added during the analysis, BrandZaps schedules another run.

Evidence: B2B audience Material
View company profile →
In keyword research

From website and material texts, BrandZaps derives real questions your audience asks and checks them against real search data. The material widens the view beyond your existing website.

Question from material checked against search data
View keyword research →
In the Project Brain

The stored material becomes part of the knowledge context. Together with profile, brand, website, topics and decisions, it extends the Project Brain.

Knowledge context extended
View Project Brain →

Bring in knowledge without filling the Project Chat with third-party text.

The Project Chat deliberately does not read the wording of uploaded files. This is not a missing feature but a security decision: third-party text can contain instructions that are slipped to an assistant.

Material is processed in the designated analyses – but not taken into the Project Chat's context as arbitrary wording.

View Project Chat
No wording in the chat context

Uploaded files do not reach the assistant as running text.

No image analysis

Images in files are not analysed and not silently treated as read.

Processing only in designated analyses

The knowledge base makes structured company knowledge available to the features designed for it.

A shared material base for the project.

The knowledge base connects previous work with the current BrandZaps workflow. The result is not a knowledge database for its own sake, but a source analyses can draw on instead of gathering dust in folders.

Notes record decisions
Files bring in existing analyses
Public research adds dated sources
Monthly folders show when knowledge was created
Profile, research and Brain draw on it
Origin stays recognisable as material

What your company already knows stays part of the project.

Notes, files and public research end up in one place – and keep working where BrandZaps prepares decisions.