Proposals grounded in real history
What price we quoted this client, what the last contract agreed, which objections came up in similar deals. The agent answers with the previous proposal attached, instead of sending you hunting through folders.
Digital Brain: the private AI that knows your company from the inside
Company knowledge stops living in two or three people heads. Contracts, procedures and history become searchable in plain language, connected to the ERP and internal systems, with access defined by profile.
What a Digital Brain is
It is a private AI layer that works on your company knowledge. Instead of answering from what it learned on the internet, it answers from your contracts, procedures, past proposals, ERP records and client history. Anyone on the team asks in plain language and gets an answer with the source in view.
The difference from a generic assistant is not the language model, it is the access. The model is the same; what changes is having the real business context, permissions by profile and a record of who consulted what.
The three components
- Private knowledge base: documents, databases and history indexed on infrastructure your company controls, never leaving for public models.
- Agents per department: each area has its own, with access only to the information it needs and the rules of the process it serves.
- Governance: permissions by profile, a cited source on every answer, and usage logging for internal audit.
The problem it solves: the knowledge is nowhere
In most companies, what makes the operation work is not written down. It is in the heads of whoever has been there longest. When that person leaves, falls ill or changes roles, the operation slows down and nobody can explain exactly why. What is written is scattered across folders, emails and systems that do not talk to each other, and nobody has time to organise it.
Signs you are at this point
- There are questions only two or three people can answer, and everyone knows who they are.
- New hires take months to become autonomous, and their training depends on whoever happens to be free.
- The same proposal or procedure gets rebuilt from scratch because nobody can find the previous version.
- The team has started pasting company information into personal AI tools, with nobody controlling what leaves.
What the team does with it day to day
It is not an internal encyclopedia nobody opens. These are concrete questions, asked at the moment the work happens.
Procedures without depending on who knows
How this exception is handled, what steps this process has, what to do when a supplier fails. Answers come from your manuals and case history, not from generic best practice.
Onboarding new people
New joiners have somewhere to ask without interrupting colleagues. Repeated questions stop consuming senior time, and the questions that appear show where the documentation has gaps.
Context before deciding
What projects of this type we have done, what they actually cost, what went wrong. Information that exists but that nobody can assemble in time for the decision.
On top of what already exists
Nothing needs organising before you start. If everything were organised you probably would not need this. The work is precisely handling scattered information and connecting it to where the work happens.
- Portuguese business software: Primavera, PHC, Sage, Moloni and legacy systems with no API
- Document folders, SharePoint, Google Drive and network shares
- Email, Microsoft 365 and Google Workspace
- CRM and internal databases, including on-premise installations
- Internal chat, so the team can ask where they already work
Privacy, the question that stops the purchase
We deal with this at the start, not at the end. Infrastructure stays in the European Union, in accounts your company controls, and documents are not used to train public models.
- Permissions by profile: whoever cannot see a contract in the system does not see it here either.
- Every answer cites its source document, so anyone can confirm it.
- There is a record of who asked what and with which sources, for internal audit.
- Sensitive documents can be excluded from indexing by folder or by tag.
- We sign a confidentiality agreement by default, before any technical conversation.
How it is implemented
Start with one department and a small set of sources. Extending later is easy; starting with everything at once rarely ends well.
- 01
Diagnosis
30 to 45 minute session, freeWe map where knowledge lives today, who depends on whom and which questions keep repeating. You get a written view of where it makes sense to start.
- 02
Design and permissions
2 to 3 weeksMap of sources to index, permission model by profile and a definition of what stays out. You end up with scope, timeline and investment fixed.
- 03
First department
6 to 8 weeksKnowledge base in production for one area, with a group of real users. We test with the questions they actually ask, not with demo cases.
- 04
Rollout
Depends on scopeWith the first department using it and measured, more sources and areas come in. This is where it starts to make sense to talk about agents that take action.
Investment
Reference bands upfront, to save both sides time. The exact figure is fixed in the proposal, after the diagnosis, and does not change midway.
Core
€25,000 to €45,000
Private knowledge base, two to three department agents, ERP or CRM integration and access control by profile. 8 to 12 weeks.
Platform
From €45,000
Multiple departments, agents that take action in your systems, and ongoing supported evolution. 12 to 24 weeks.
Maintenance
On proposal
The first 3 months of support are included. After that, an evolution contract that fits, never as a dependency.
Frequently asked questions about the Digital Brain
ChatGPT answers well from what exists on the internet, but it does not know your clients, does not open your ERP, does not respect your permissions and leaves no record. It helps one person write faster. The Digital Brain answers from your documents and systems, with the source in view and access controlled by profile.
No. If you had everything organised you probably would not need this. Part of the work is handling information scattered across folders, emails and systems that do not talk to each other. What we need to know is where the information lives and who is allowed to see it.
No. Infrastructure stays in the European Union, in accounts your company controls, and documents are not used to train public models. We define in writing which sources are indexed and which stay out.
Every answer shows the document it came from, so a mistake is verifiable rather than invisible. When the information is not in the indexed sources, the system says it does not know instead of inventing. Before going live we test with the team real questions.
Typically 6 to 8 weeks for the first department, after 2 to 3 weeks of design. Counting from the first conversation, it is reasonable to expect around two and a half months until people are using it daily.
The code, the infrastructure and the indexed data are yours. We hand over technical documentation, access and team training. You can move maintenance to another team or bring it in-house without renegotiating anything with us.
Next step
Want to know what this would do in your operation?
Free diagnosis, no commitment. We look at your process and come back with a plan covering scope, timeline and investment. Reply within 48 business hours, NDA included by default.
- 100% free diagnosis
- No commitment to proceed
- Reply within 48 business hours
- NDA included by default
Request free diagnosis
Two minutes to tell us the essentials. We reply within 48 business hours with a first read on your case and, if it makes sense, a proposed session.
Explore other solutions
Every vertical has its own challenges. See how we approach other sectors where we have proven experience.