Knowledge sources
We select the documents and systems the assistant will answer from and check that they are current and that access rights are correct. What the assistant must not touch is defined from the start.
An assistant that knows your policies, price lists and contracts, not the internet. We implement Claude assistants for employees and customers, with your company knowledge, access rules and a clear answer on where the information came from.
Built for ISO 27001NIS2 DORA
A generic chatbot answers nicely but does not know your leave policy, what your service costs or what your contract with a customer says. A company assistant must answer from your documents, see only what the employee asking can see, and cite the source so the answer can be checked.
We build Claude assistants with a Claude Team / Enterprise workspace with projects and knowledge bases, or a custom assistant built on the API. We connect knowledge sources (policies, price lists, product descriptions, contract templates) according to existing access permissions, set the tone and limits (what the assistant must not do) and measure how many questions are answered without a person.
The scenarios we implement most often with Claude. Each one starts with a single process and is measured.
From the first conversation to a working solution, with measurement and ongoing support.
We select the documents and systems the assistant will answer from and check that they are current and that access rights are correct. What the assistant must not touch is defined from the start.
We build the assistant with a Claude Team / Enterprise workspace with projects and knowledge bases, or a custom assistant built on the API, and set the tone, language, limits and the requirement to cite the source.
A pilot group asks real questions for two weeks. We measure how many are answered correctly and without a human, and fix gaps in the knowledge.
We launch to everyone, track unanswered questions and refresh the knowledge sources every month.
No more than the employee asking could see anyway: the assistant works with existing access permissions. On commercial plans and through the API, your data is not used to train models; with Amazon Bedrock the model can run in an EU region, and access keys and logs stay under your control.
It says it does not know and points the user to a person. We set this in the rules. We review unanswered questions and add to the knowledge sources.
Yes. Typically there are separate assistants for sales, HR and customers, because knowledge sources and tone differ. We start with one.
You pay by usage (API) or per user (Claude Team / Enterprise); we help choose the cheapest combination for your workload. We give the implementation price after the first conversation.
We automate what repeats: invoices, orders, reports, contract reviews.
An agent is not a chat window but a coworker without hands: it receives a task, gathers data from your systems, takes action and reports back.
AI is most useful not in a separate window but in your systems: ERP, CRM, helpdesk, document management.
This service is delivered by specialists with valid international certifications. Their expertise is confirmed by independent organisations, not by us.
AI systems are deployed with protection against data leakage and model abuse.
The AI security programme and its controls are managed following the ISACA methodology.
Personal data is handled according to privacy by design and GDPR requirements.
We reply within 1 business day