An AI assistant that knows your business.

What is an AI assistant?

An AI assistant is software that understands and answers questions in natural language. You type what you want to know, the assistant answers in seconds. Behind it works a language model that understands, summarises and drafts text. What matters is the knowledge base. Consumer assistants are trained on public internet data. They answer general questions well, but they know neither your price list nor your holiday policy. A business AI assistant is built differently. It is connected to your company knowledge, at rinqo via a central knowledge tree built from website content and uploaded PDFs. Asked “How do we handle a complaint?”, it answers with your process, not an internet definition. Every answer carries a source citation, you see which document the information comes from. The second difference is legal. Consumer tools usually run on US infrastructure, without a data processing agreement for your business. A business assistant needs three things: EU hosting, a DPA per Art. 28 GDPR and the assurance that your data does not flow into model training. rinqo meets all three, Hetzner data centres in Germany, DPA included, no training on customer data. In short: a consumer assistant knows the internet. A business AI assistant knows your business, and can prove it.

What an AI assistant with company knowledge can do

An AI assistant with company knowledge answers the questions that otherwise travel across the team. New employees no longer ask the colleague at the next desk, they ask the assistant: how do I request leave, where is the quote template, who handles data protection. The assistant answers in seconds and names the source, say the onboarding PDF or the imported process page. That relieves experienced staff who would otherwise answer the same questions for the tenth time. Process knowledge is the second big use case. How does a complaint run, what are the steps of order intake, what applies to cancellations: the assistant knows the documented procedures and returns them in a structured way. Because all answers come from the same knowledge tree, every person gets the same current information, not three variants from three heads. The source citation makes every answer verifiable. If a document contains something wrong, you spot it immediately and correct the document, not the assistant. Over time the knowledge base gets better, not worse. The assistant understands 17+ languages with automatic detection. A Polish-speaking colleague asks in Polish and receives the answer from the same German documents.

• Onboarding: leave requests, templates, responsibilities, without interrupting colleagues
• Process knowledge: complaints, order intake, cancellations, structured and current
• Source citations: every answer with a reference to the original document
• 17+ languages: automatic detection, one body of knowledge for everyone

Securing knowledge before it walks out the door

An AI assistant secures knowledge that otherwise leaves the company with people. Three situations hit every SMB: retirement, resignation, cover. Retirement is the clearest case. When the long-serving colleague leaves, edge cases, customer histories and practical know-how often leave with her. The remedy is unspectacular: document the knowledge, load it into the knowledge tree as PDFs, test it. From then on the assistant answers questions that previously only one person could answer, with a source citation on the document. With a resignation the time pressure is higher, the principle stays the same. Handover documents get written anyway, in the knowledge tree they become queryable instead of filed away. A folder on a network drive answers no questions, an AI assistant does. The third case is the most common: cover during sickness or holidays. The person covering accesses the same knowledge as the person covered, instead of digging through inboxes and folders. Questions like “What terms does customer X have?” or “How does month-end invoicing run?” are answered from the documented material. The side effect applies to all three cases: documentation gains a visible purpose. What sits in the knowledge tree gets queried daily, gaps show up and get closed. Knowledge that lives in the knowledge tree doesn't hand in its notice.

Data protection for AI assistants

A business AI assistant must meet three data protection requirements: EU hosting, a data processing agreement per Art. 28 GDPR and no training on your data. rinqo meets all three. Hosting runs in Hetzner data centres in Falkenstein and Nuremberg, certified to ISO 27001:2022. Language model processing runs contractually guaranteed within the EU. No model is trained on customer data, your documents remain your documents. A purely European fallback provider stands by as a failover, as an opt-in, you decide whether it is active. The DPA per Art. 28 GDPR is provided. On top comes the EU AI Act 2024/1689. Art. 4 (AI literacy) has applied since 2 February 2025, whoever uses AI in the company promotes AI literacy among the staff who work with it. Art. 50 (transparency obligation) applies from 2 August 2026: people must be informed when they interact with an AI. Fines for violations: up to €15 million or 3% of global annual revenue, the lower amount for SMEs. rinqo activates the disclosure by default. In chat and email it cannot be switched off; on the phone it is part of the greeting, whose wording you adjust. For you this means: the compliance foundations are part of the platform, not your project. You maintain knowledge, not legal texts.

Setting up an AI assistant: the process

Setting up your first AI assistant runs by us at rinqo. The flow: create an account, enter your website address, rinqo reads the content automatically and builds the knowledge tree from it. Then you upload PDFs, price lists, onboarding material, process documents. Then you test: ask the ten questions that come up most often in your team and check the answers including sources. If something is missing, you add the relevant document, the next answer lands. The knowledge tree is shared. What you maintain for the Personal Agent also powers the phone agent, the chat agent on your website and the email agent in your inbox. You build the knowledge base once and supply every channel. You set up calendars (Google, Outlook or CalDAV), the REST API and webhooks yourself. CRM and accounting (HubSpot, sevdesk, lexoffice) or project tools (Linear, Jira) we connect on request. No minimum term, no per-seat licence, set up by us from the start.