Artificial intelligence and physics technology firm SandboxAQ announced the launch of Switch, an open-source software solution designed to integrate autonomous AI agents into existing team collaboration platforms. Available immediately on GitHub at no cost under the Apache 2.0 license with Commons Clause, Switch enables organizations to establish shared digital workspaces across Slack, Microsoft Teams, Discord, and other enterprise messaging applications allowing human workers and AI agents to collaborate seamlessly within unified channels.
While developers continue to build highly capable AI agents, operationalizing these models across multi-disciplinary teams frequently requires custom integrations, manual context sharing, and fragmented tool switching. Switch eliminates this coordination friction by establishing model- and platform-agnostic environments. Utilizing Model Context Protocol (MCP), standard APIs, and flexible adapters, the software allows engineering teams to connect agents built on diverse frameworks such as OpenAI, Claude Code, Google ADK, and LangChain without locking business workflows into a single vendor ecosystem.
“Every team we talk to has capable agents trapped in separate tools, and a coordination tax eating the productivity those agents were supposed to deliver,” said Mohammed Aboul-Magd, General Manager, SandboxAQ. “Switch pulls agents into the channels where teams already work, where knowledge compounds instead of being relearned with every task. If you can message a colleague, you can put an agent to work.”
Also Read: DXC Technology and Primary Unveil AI-Native Zero Trust Platform to Secure Enterprise AI
Preserving Context and Eliminating Cross-Tool Friction
Engineered to transform standard chat channels into dedicated operational rooms, Switch maintains project timelines, resource files, and conversational history directly within the workspace. As specialists or specialized agents join, leave, or hand off tasks, project context remains intact—preventing information resets and accelerating output across complex enterprise workflows.
Key technical capabilities of Switch include:
Open-Source Architecture: Rapid deployment models for Quick Start and Team editions, allowing enterprise IT departments to deploy without vendor lock-in.
Framework Interoperability: Native compatibility for custom agents created via Claude Code, Google ADK, LangChain, OpenAI, and other prominent development suites.
Unified Channel Workspace: Embeds autonomous agents directly into existing corporate messaging systems, including Slack, Microsoft Teams, and Discord.
Persistent Context Preservation: Retains historical channel context, ongoing documentation, and operational rules through every task hand-off.
Multi-Room Agent Scalability: Permits a single AI agent to operate simultaneously across multiple project channels while preserving unique context boundaries for each room.
Real-World Multi-Disciplinary Applications
SandboxAQ’s internal engineering, design, and marketing teams utilized Switch throughout its own software development lifecycle. By structuring workflows around dynamic task channels, cross-functional teams applied specialized agents across distinct enterprise scenarios:
Incident Response: Security and DevOps engineering teams pull specialized diagnostic agents into live incident channels, instantly providing models with real-time system timelines without manual briefings.
Marketing Campaign Execution: Content teams staff campaign rooms with dedicated research, copywriting, and design agents, building assets sequentially within a shared history.
Legal Drafting & Review: Legal departments integrate contract drafting and compliance audit agents that maintain full deal history, compressing contract review cycles.
Ecosystem Expansion via Flint AI
Switch forms a core component of Flint AI, SandboxAQ’s dedicated suite of enterprise AI agent products. The software extends the local agent testing capabilities offered by the free Flint CLI tool, providing a direct pathway from individual model evaluation to full-scale team deployment inside enterprise environments.