01What your company knows
Knowledge
The AI knows your documents, policies and minutes and answers with the source. Documents and the knowledge base stay in your own environment.
Connected sources
- Documents
- Policies
- Minutes
- File storage
HABB Core connects your knowledge and your systems into one point of contact for your whole team. The AI answers questions, fetches data from ERP and CRM and carries out tasks with your approval. It runs in your own rack or in your private cloud.
Data stays in-house
Only what you approve goes out
No change of systems
ERP, CRM and file storage stay as they are
Fixed, predictable costs
The language model runs locally, with no token costs
How HABB Core works
Your team asks where it already works. HABB Core decides where the answer comes from and passes the request on itself.
HABB Core
Running in your own rack
You, Teams
What does our expense policy say about overnight stays?
HABB CoreExample
Overnight stays are reimbursed up to the maximum rate per night. Hand in the hotel invoice with your expense claim.
From your knowledge base · in-house
You, Browser
What is the status of order 4711?
HABB CoreExample
Here is the current status from your ERP:
Read straight from the ERP · nothing changed
You, Mobile
Find a slot with the project team next week.
HABB CoreExample
Scheduling is handled by Microsoft Copilot in your organisation. I have passed the request on. The suggestions for the project team:
Handed to Copilot · logged
You, Teams
What does the new change in the law mean for us?
HABB CoreExample
For this I bring in a cloud AI. Only your approved question goes out – your documents stay in-house.
Logged · tokens are only used here
Example questions
Knowledge base and language model run in your environment – on your own hardware or in a private cloud of your choice.
Answers follow permissions
The AI answers with the rights of the person asking.
Every hand-over is logged
When the AI passes a task to an assistant or a cloud AI, that is logged.
In the tools you know
In Teams, in the browser or on a mobile – no new tool to learn.
HABB Core in your IT
HABB Core is built into the IT you already run. Your systems, tools and IT partners stay as they are. The AI connects the systems that do not talk to each other today.
01What your company knows
The AI knows your documents, policies and minutes and answers with the source. Documents and the knowledge base stay in your own environment.
Connected sources
02What your systems know
Through open interfaces (APIs) the AI reads directly from your ERP, CRM and business applications. Tasks such as finding a meeting slot are handed to existing assistants such as Microsoft Copilot.
Connected systems
03Controls what leaves the building
If a question needs a cloud AI, only the approved request leaves the building. Every delegation is logged. Token costs arise only then.
Controlled by
Each person only sees what they are already allowed to see today.
The AI starts read-only. Actions in your systems happen only with approval from a person, and are logged.
Open building blocks. Configuration, learned procedures (skills) and knowledge belong to you and can be exported at any time.
Where it starts
A workshop with management and key people – the process first, then the technology. You receive prioritised use cases and a recommendation: your own hardware or a private cloud, with a cost comparison over three years.
Each stage has its own result. After each stage you decide whether to continue.
Stage 1
Knowledge
6–10 weeks
The AI answers questions from your company knowledge – with the source.
Stage 2
Read systems
+4–8 weeks
ERP, CRM and business applications are connected – read-only at first.
Stage 3
Act
+6–12 weeks
Actions in your systems – always approved by a person, and logged.
Stage 4
Learn
ongoing
Verified solutions are stored internally as skills.
Hosting and costs
You decide where HABB Core runs. Either way, documents and the knowledge base stay in your own environment.
HABB Core runs in a rack or server cabinet on your premises.
HABB Core runs in a private cloud of your choice.
Which option suits you is settled in the AI and process check – with a cost comparison over three years.
The language model runs locally. That keeps the costs fixed and predictable – however intensively your team uses the AI.
Tokens are only charged when a question is deliberately delegated to a cloud AI – through the gateway, with a budget and a log.
Illustrative example
≈ 20hours a week
CHF 6'900a month
Employees × minutes ÷ 60 = hours a week. Hours × internal cost × 52 weeks ÷ 12 months = amount a month, rounded to CHF 100.
Example values, not a promised saving. Which use cases pay off for you is what the AI and process check shows.
Approach
In the AI and process check we go through your processes with management and key people. For each process we ask the same four questions.
01Input
What comes in?
02Output
What should come out at the end?
03Steps
Which steps lie in between?
04Person or machine
Which step stays with a person, which one does the machine take on?
Your question is not listed? Write to us.
No. Documents and the knowledge base stay in your own environment – in your own rack or your private cloud. At most an explicitly approved request goes out to a cloud AI, through the gateway with data filter, log and budget. Every delegation is logged.
No. HABB Core is built into the IT you already run. Your systems, tools and IT partners stay as they are. ERP, CRM and business applications are connected through open interfaces (APIs).
Your existing permissions stay in place. Each person only sees what they are already allowed to see today.
It starts read-only. Actions in your systems only come with the «Act» stage – always approved by a person, and logged.
The language model runs locally, so there are no token costs: the costs are fixed and predictable, however intensively the AI is used. Tokens are only charged when a question is deliberately delegated to a cloud AI. The AI and process check gives you the cost comparison between your own hardware and a private cloud over three years.
In the tools they already know: in Teams, in the browser or on their phone. No new tool is needed.
It runs in stages: Knowledge (6–10 weeks), Read systems (+4–8 weeks), Act (+6–12 weeks) and Learn (ongoing). Each stage has its own result. After each stage you decide whether to continue.
No. We build on open building blocks. Configuration, learned procedures (skills) and knowledge belong to you and can be exported at any time.
With the AI and process check: A workshop with management and key people – the process first, then the technology. You receive prioritised use cases and a recommendation: your own hardware or a private cloud, with a cost comparison over three years. At a fixed price, credited towards follow-on work and with no obligation.
One workshop with management and key people. Afterwards you know which use cases pay off and whether your own hardware or a private cloud suits you better.