# From chatbot to co-worker: searchit brings agentic AI to companies Press release of 23 September 2026: the searchit AI Agent makes a company's data holdings available to the AI as its memory and carries out tasks like a digital co-worker. *Vienna, 23 September 2026* AI agent works with company knowledge – at a predictable price **(Vienna, 23 September 2026)** Most companies have so far known artificial intelligence as a chat window: you ask a question and get an answer. But this AI knows nothing about your own organisation, and the actual work stays with people. The leading enterprise search provider “searchit” goes one step further with the searchit AI Agent. It is one of the first agentic AI solutions that makes a company's entire data holdings available to the AI as its memory, at a price that is predictable even for mid-sized businesses. The agent has access to file shares, mailboxes and internal systems, understands what is in them and acts on it: it compares contracts, drafts letters, creates Excel analyses and PDF reports and sends them when asked to. A passive chatbot thus becomes a digital co-worker. ## What AI lacks is not intelligence but a memory A language model can phrase, summarise and draw conclusions. But it does not know which contracts a company has with a customer, which invoices are outstanding or what was in the last quote. “What AI lacks is not a better model, it lacks a memory,” says Ing. Christoph Wendl, CEO of searchit. “Companies have had this memory for a long time: in their documents, mailboxes and drives. searchit makes it accessible, and the agent works with it.” To do this, searchit reads every document with AI, summarises it and recognises people, companies, places and amounts in it. The result is a knowledge graph that shows how customers, contracts, invoices and contacts are connected, even if the same company is spelled differently in two documents. More than 100 sources can be connected: file servers, SharePoint, Teams, email mailboxes, specialist databases and web sources. ## A co-worker who gets things done The searchit AI Agent does not just answer questions, it carries out tasks. Around sixty tools are available to it for this. If, for example, a sales employee asks it to compare two framework agreements with customers, it finds the contracts via the knowledge graph, reads them in full and saves an Excel table for her with term, services, payment terms and discounts. If accounting needs a payment reminder, it reads the open items list and previous reminders and creates a Word draft from them, which is only sent when a person instructs it to. And if the task is to summarise a customer on one page and send it to sales, the agent gathers contracts, open items and contacts, creates a customer sheet as a PDF and sends it by email. In addition, the agent sorts mailboxes, schedules appointments, tags entire document collections, redacts personal data and even operates web applications that have no interface. Recurring tasks such as “show me the new tenders every Monday” run on a schedule. It can be reached at the workplace, in the searchit client and on the move via messengers such as Signal or WhatsApp. ## A week becomes a day How much manual work can be handed over is shown by a customer project from 2026. A sales manager at a construction company with around 200 employees is preparing a bid for a tender: reviewing criteria and standards, evaluating existing offers and service catalogues, adopting clauses from the general terms and conditions and involving subcontractors by email. This used to take around a week; with the searchit AI Agent it takes around one person-day. “The biggest item is rarely search time,” explains Wendl. “It is the work that people do by hand today because no specialist system holds all the necessary information in one place: gathering documents from several systems, rebuilding the same analysis every week, retyping information from a web application. This is exactly the work the agent takes over.” ## With the rights of the person, not with a master key An agent that can change files and send emails needs clear limits. The searchit AI Agent therefore always works under the login of the employee who instructs it. It sees and does only what this person is also allowed to do in the company's system. Each tool group can be enabled or blocked per role, and every step is logged with user, time and result. Password and PIN fields are removed before any text is sent to the language model. The data stays in an installation that belongs only to the respective company. The customer decides where the agent runs: at the workplace under Windows, Linux or macOS, centrally on a Windows terminal server or in the cloud on its own virtual machine. The EU AI Act has also been considered: every AI-generated field carries model, version, time and prompt ID, machine-readable and also in the export. More about the searchit AI Agent: [searchit-agent.ai](https://searchit-agent.ai/) ## About searchit searchit turns an organisation's knowledge into the memory of its AI. The searchit AI Agent knows your data holdings (emails, documents, etc.) and works with them. This makes the agent a digital co-worker. The technology builds on searchit's long experience as an enterprise search platform used by numerous customers such as Samsung, TÜV, the Technisches Museum Wien and the Bavarian digital authority. searchit is made by Iphos IT Solutions GmbH, based in Vienna. Its in-house AI lab continuously develops the platform's AI functions. Ing. Christoph Wendl, together with Lyubomir Ivanov, runs the company, founded in 1998, as Chief Executive Officer. ## Media contact - Ing. Christoph Wendl, Managing Director, Iphos IT Solutions GmbH - Khekgasse 35, 1230 Vienna, Austria - Phone: +43 1 869 84 00 - Email: [info@searchit.cc](mailto:info@searchit.cc) - [Release as PDF (German)](https://searchit-agent.ai/pr/vom-chatbot-zum-mitarbeiter-searchit-bringt-agentic-ai-in-unternehmen/20260923-presseaussendung-searchit-ai-agent.pdf) --- - searchit AI Agent: This version is generated from the page's language files (scripts/gen-unternehmen.py). The page itself is authoritative.