Fathym launches governed industrial data-to-AI platform

By AI, Created 11:35 AM UTC, May 20, 2026, /AGP/ – Fathym has launched Open Industrial, a platform that connects industrial and life sciences operational data to AI tools while adding governance, approval workflows and audit trails from the first query. The move targets regulated teams that want to use tools like Claude Code and Codex without rebuilding fragile custom integrations.

Why it matters: - Industrial and life sciences teams are adopting AI faster across engineering and scientific workflows. - Open Industrial is designed to let those teams use AI on live operational data without giving up governance, approval controls or auditability. - The platform is aimed at regulated environments where data integrity, chain of custody and traceability matter.

What happened: - Fathym announced the launch of Open Industrial, a governed data-to-AI platform for industrial and life sciences operations. - The platform is available immediately. - A self-serve tier is available at openindustrial.co. - Enterprise engagements are offered as fixed-scope pilots deployed in the customer’s own Azure tenant.

The details: - Open Industrial connects IoT devices, REST APIs, SQL databases, laboratory information management systems and process historians to AI tools teams already use. - The platform supports configuration through a visual workspace, reducing the need for custom integration code. - Azi, the built-in AI assistant, proposes queries and workflow logic based on the team’s environment. - Human approval is required before any query executes against a connected system. - Every query and approval is exposed through a Model Context Protocol server. - MCP-compatible tools can get read and write access to the governed data layer. - Engineers using Claude Code, Codex or other MCP-compatible clients can query live operational data or propose new workflows, nodes and integrations. - Structured metadata returned in every API response identifies who approved each query or change, when it happened and the reasoning behind it. - The platform uses open protocols and standard HTTP endpoints throughout. - All components are deployed within the customer’s own Azure tenant. - Customers retain ownership of every data connection, query and API.

Between the lines: - The platform is built around a vendor-neutral idea: companies can switch AI tools without rebuilding the underlying data stack. - That matters for regulated organizations that want AI access but cannot accept opaque, untracked system changes. - The proposal-before-execution workflow is the core control point that separates Open Industrial from a simple data connector. - Fathym is also positioning the platform around a pharmaceutical sample management use case, where auditability and chain of custody are especially strict.

What’s next: - Fathym is developing enterprise use cases in pharmaceutical sample management. - The company is pitching fixed-scope pilots for enterprise customers that want the platform inside their own Azure environment. - The same governed architecture used in that use case is intended to carry across other deployments.

The bottom line: - Open Industrial is Fathym’s attempt to make operational AI usable in regulated settings without sacrificing control, traceability or data ownership. - Matt Smith, CEO of Fathym, said the goal is to let teams use any AI tool against their own operational data with a complete audit trail from the first session.