Oulmanak launches a cloud agent with cumulative wiki memory and an always-on computer

Operating models for AI agents inside enterprises are moving toward overcoming the structural limitations of chatbots and traditional software integrations, which are limited to fetching data in real time and losing context once the session ends. In this context, Oulmanak, as part of the summer 2026 batch of the Way Combinator accelerator, unveiled a platform for an AI agent that runs on an independent, always-on cloud computer, with its own browser, files and login credentials for various enterprise work tools, allowing it to perform tasks in the background even after the user shuts down their personal computer.

The platform relies on a dual architecture that combines automated execution with what it calls the organization’s second brain, a wiki-style knowledge base that updates itself continuously and accumulatively from daily activities. The agent connects to workplace applications such as messaging channels, email and meeting notes, extracting decisions, project updates and client records to merge them into unified documentation pages.This cumulative wiki serves as the mandatory reference that the agent reads before starting any task.This gives it prior awareness of agreements and transaction history without needing to be re-provided context with each new request.

The platform allows task delegation via text messages or collaborative work channels, with the agent tracking them and sending a notification upon completion, and it can automatically initiate handling of operational gaps. Execution scenarios range from monitoring fault reports in support channels and opening documented technical tickets with reproduction steps and screenshots, to retrieving agreed pricing terms with clients, or collecting digital invoices and attaching them to banking transactions.Sensitive operations such as payments or logins are subject to direct human oversight.The agent requests prior approval or allows the user to take instant control of the cloud browser.

At the data-management level, the platform separates individual connected accounts from the general corporate context, ensuring that only summaries and useful decisions flow into the shared wiki without exposing personal mailboxes, while allowing manual editing of wiki pages and review of links to each entry’s source. Users also retain the ability to monitor execution step by step and can revoke access rights to any connected tool at any time.

This operational model has a practical impact on startups and tech teams in the Gulf, Egypt and the Levant, where fast-moving work environments suffer from fragmented knowledge across multiple platforms and loss of decision context as tasks cascade.Relying on always-on agents with a unified memory changes the productivity and corporate knowledge-management equation.It shifts teams from wasting time on manual documentation and routine task tracking to managing agent permissions and reviewing their outputs, as well as cementing skills in auditing documentary memory and controlling data flow to ensure compliance with security standards adopted by regional organizations.