Process Street launches MCP server for AI access to compliance workflows

Process Street has launched a hosted Model Context Protocol server that lets AI assistants reach live workflows, tasks and data sets inside the platform without custom integrations. The move extends access to Claude, ChatGPT and Microsoft Copilot Studio and keeps permissions tied to existing user roles.

Why it matters: - Process Street is making compliance and operational workflows usable from AI assistants without building custom API integrations first. - The change could speed up everyday work like reviews, task assignment, reporting and data searches for teams already using Claude, ChatGPT or Microsoft Copilot Studio. - The server keeps access limited to each user’s existing Process Street permissions, which matters for auditability and control.

What happened: - Process Street launched a hosted Model Context Protocol server at mcp.process.st. - The server is available on all Process Street plans. - The server connects AI assistants and other applications to workflows, workflow runs, tasks, users and data sets outside the Process Street platform. - The server works with Claude, ChatGPT on Pro, Team, Business, Enterprise and Edu plans, Microsoft Copilot Studio and any application that supports Model Context Protocol. - One-click connectors are available for Claude, ChatGPT and Microsoft Copilot Studio.

The details: - Model Context Protocol is an open standard for connecting AI assistants to external systems. - The new server supports most Process Street API endpoints as tools. - Connected assistants can create and manage workflow runs, assign tasks, search data sets, generate reports and summaries, and analyze operational data through conversation. - Process Street’s documentation gives examples that include launching a Quarterly Access Review workflow run and assigning it to a compliance manager if needed. - Another example shows an assistant finding the most recent Client Onboarding workflow runs and identifying bottlenecks from incomplete tasks. - A third example shows searching data set records for active vendors and flagging those not reviewed in the past six months. - Admins, Builders and Users can use the MCP server, but each role can only reach resources allowed by existing Process Street permissions. - Process Street is not introducing a new permission model for the MCP server. - No data is moved out of Process Street to make it available through the server. - Organizations with SAML SSO enabled are not supported at this time. - Documentation and connection instructions are available at the MCP server documentation.

Between the lines: - The launch reflects a broader shift toward AI assistants becoming the front end for day-to-day operations software. - Process Street is betting that the winning approach is not looser access, but making existing procedures legible to assistants while preserving controls. - CEO and co-founder Vinay Patankar said the key gap is operational context, not model quality, and that customers already documented their processes.

What’s next: - Process Street says organizations can now connect supported assistants directly to live compliance workflows under existing permissions. - The practical test will be how quickly teams use the server for routine operational work and how broadly other MCP-capable applications adopt the integration. - Process Street says its ISO 27001 and SOC 2 Type II certifications, along with HIPAA and GDPR compliance, remain in place.

The bottom line: - Process Street is trying to make AI assistants useful inside regulated operations without weakening the security and permission controls that those teams depend on.