Blog · Published · Updated
Budgeting for Company Brain content upkeep
Short answer
Keeping knowledge current in a company AI assistant has two costs. The first is the time of the people in the company who correct documents after changes. The second is the contractor's care for the application, priced by scope. In a scenario with 32 updates a month the team spends 17.6 hours instead of 24. A narrow set of frequently used documents with a named owner is the cheapest to keep up.
How we workDetails ↓
- 01Call30 minutes with an engineer, free
- 02PrototypeA working system on your process and data, free
- 03ImplementationTarget and fixed price in the contract; 50% at signing
- 04Acceptance50% once the agreed target is met
- net per application, fixed price before contract
- from €3,000
- from signing the contract to acceptance
- 4–10 weeks
- free prototype on your documents
- €0
We belong to the OpenAI and Anthropic partner programmes and learn through their official courses. What this means ↓
Author: Syntalith · 4 min read
Statistics
Company knowledge and AI: statistics 2026
17.6 h
a month in place of 24 h for 32 guide updates, in a scenario with a tool that prepares the changes and a person who checks them.
62%
According to Microsoft's Work Trend Index (2023), 62% of people struggle with spending too much time searching for information in their workday.
15.1%
According to Eurostat (2025), 15.1% of people aged 16 to 74 in the EU used generative AI tools for work.
48%
According to the European Commission (2026), 48% of EU SMEs and mid-caps name a lack of expertise as a barrier to AI.
Source: European Commission, Flash Eurobarometer on AI (2026) ↗
20.0%
According to Eurostat (2025), 20.0% of EU companies with at least 10 employees used AI.
Figures checked: 22 October 2026
The two costs of keeping knowledge current
In an assistant's budget the subscription and the care for the application are easy to find. Harder to see is the time of the person who corrects a guide after a product change and checks the new answers. If that work has no place in the team's budget, the knowledge base grows faster than anyone can keep it up.
The company team's time
hours a month
Correcting content after a process change, an expert confirming it makes sense, settling conflicting versions.
The contractor's care
by scope
The new version reaches the answers, source links repaired, application errors fixed, monitoring.
Models and hosting
by usage
Search and answers grow with the number of questions employees ask.
How long an update takes: a scenario
Let us count the update of a guide after a product change. Today an editor does everything by hand. With a tool the change is prepared faster, but the human check stays.
- Updates a month
- 32
- By hand, the whole update
- 45 min
- With the tool: preparation
- 15 min
- Check by an employee
- 15 min
- Buffer for questions and corrections
- 3 min
- Today
- 32 × 45 min = 24 h
- With the tool
- 32 × 33 min = 17.6 h
- Model: 20,000 + 3,000 tokens per update, Claude Sonnet 5.5
- €2
- Value of the time given back
- 6.4 h × €34.90 = €223
Left after the model cost
€221 a month
All times and token counts are the scenario's assumptions. €34.90 an hour is the EU average hourly labour cost for 2025 according to Eurostat. The model at Anthropic's rates, at the ECB rate of 1.1205 USD per euro on 22 October 2026. Time your own updates before you set a budget.
Time given back means the team is free for other work. It becomes a saving when it changes spending, for example when it cuts overtime.
Split the tasks in the contract
A report that says „the assistant answers wrongly” can mean an unclear document or an application error. Agree in advance who handles which case.
The process has changed and the document is outdated
the document's owner in the company
team time
The new version does not show up in answers
the contractor
in the care plan
Two documents say different things
the process expert decides
team time
An answer misquotes a correct document
the contractor
in the care plan
A department wants to add a new source
new scope
priced separately
Allow for periods with more changes, for example the launch of a new offer. A fixed care scope may not be enough then. The buyer should know how to order extra work and what comes first when several departments report updates at once.
A smaller source is often a better scope
A library of frequently used instructions deserves a different work rhythm from an archive of presentations. Base the choice on employees' questions, how often things change and what happens when someone uses wrong content. Do not keep a wide collection up to date only because it can be connected.
Confluence analytics shows views, content updates and search terms, among other things. This data helps you choose an area. It does not tell you how much work a correction needs.
- Who owns each new source
- How often these documents change
- How long the last real update took and who took part in it
- Whether an existing editorial process can be linked to the assistant without building a second one
How to price the contractor's care
We price care for the application by scope, separately from the build. We write the scope into the contract, so both sides know what is in the price and what counts as new work.
- Monitoring and the response time to an error
- How many new document versions a month the fixed fee covers
- How to order a new source and how it is priced
- Who has access to the code, the keys and the documentation
- How often you review answers that employees marked as wrong
Building the assistant is an AI app from €3,000 net, typically up to €7,000. After launch the model cost grows with the number of questions. In the scenario above the model costs €2 a month, so people's time is the largest item of upkeep.
How to start the budget conversation
Tell us how an important instruction was last corrected and who took part. From that we will describe the limits of responsibility and the scope options before you pay for an extension. The cost of building the assistant itself is covered in how much an AI assistant for documents costs, and the scope of care on the AI system maintenance page.
Worked example · Scenario
How we count the payback: the team's questions to documents
In the free process review we also count the upkeep: how many documents change in a month and who corrects them.
€3,723
a month for finding answers to 800 questions in documents
Payback: 1 mo.
Our assumptions are 8 minutes to find an answer in documents or ask a colleague, and checking in 15% of the old time. €34.90 an hour is the EU average hourly labour cost for 2025 according to Eurostat, for companies with 10 or more employees. Model cost: Claude Sonnet 5.5, about 6,000 + 500 tokens per answer from documents, at the ECB rate of 1.1205 USD per euro on 22 October 2026.
See the full calculation →
