Show HN: I built an open-source alternative to Claude Cowork (github.com)

Valmis, an open-source cloud application for deploying AI agents, has been launched on Product Hunt. This innovative platform enables users to create a fleet of secure AI agents that can integrate with over 100 business and productivity applications, including Google Workspace, Slack, and Salesforce. The agents operate in isolated containers, which means they cannot directly access sensitive information like API credentials, addressing significant security concerns prevalent in existing tools like OpenClaw. Instead, a proxy system allows agents to request data through the host machine, enhancing security while maintaining functionality. The significance of Valmis lies in its ability to automate complex workflows through a user-friendly interface, where users can create multi-step processes triggered by various events. Each agent has a sophisticated memory system that retains information across sessions, enabling continuous learning and improved interactions over time. Additionally, Valmis supports a wide range of language model providers, making it a versatile tool for businesses seeking to harness AI in a secure and efficient manner. The project’s commitment to privacy, data security, and customization is likely to resonate within the AI/ML community, particularly for organizations that prioritize safe and robust deployment of AI technologies.