Seekr joins MACH Alliance to boost AI agent traceability

Seekr has joined the MACH Alliance's Agent Ecosystem as a Collaborator, adding its explainability expertise to the group's work on AI agents that operate across vendors.

It will contribute experience in data provenance, lineage, explainability and sovereignty to the alliance's work on open standards and reference architectures. Its focus is on helping organisations trace how decisions are made when multiple AI agents are linked in production systems.

The issue has become more prominent as companies move beyond single AI assistants in test settings and begin connecting agents from different suppliers to legacy and modern software environments. In those arrangements, a decision may pass through several systems, making it harder to identify which source data informed the result, how that data changed and why a final action was taken.

That creates a particular challenge for regulated sectors and government users, where organisations may need to justify an automated decision to auditors, regulators or courts. Seekr's entry into the MACH Alliance's Agent Ecosystem reflects growing industry attention to governance and traceability as interoperability between AI systems expands.

Under the collaboration, Seekr will take part in the Enterprise Agent Architecture working group. It will also work with other members on material applying MACH principles to AI governance and share lessons from deployments in regulated and government environments.

The MACH Alliance is a vendor-neutral non-profit group focused on open, composable enterprise technology. It says it has more than 116 member companies, while its Agent Ecosystem, launched in late 2025, is backed by more than 70 supporters.

The initiative is intended to develop reference architectures, operating patterns and enterprise-scale proofs for AI agents that work through open standards rather than within a single vendor platform. The aim is to give businesses a framework for connecting specialised agents from different providers while avoiding lock-in to one system.

Tracing decisions

The alliance's work comes as enterprises test whether groups of AI agents can handle more complex tasks than a single assistant. But efforts to make those agents communicate with one another have exposed a second problem: whether companies can reconstruct how a decision was reached once multiple tools and providers are involved.

That concern has sharpened as businesses explore agent-based systems for workflows spanning customer service, operations, compliance and internal decision-making. In such cases, a flawed action may force teams to retrace an entire sequence across several agents to identify the point of failure and explain the outcome.

Rob Clark, President of Seekr, said the company sees this as the next hurdle for enterprise adoption. "Interoperability gets agents talking to each other. It doesn't tell you why they reached a decision, or which source drove it, which becomes even more complex with agents from multiple vendors," Clark said. "We've spent years helping government and regulated customers answer that question with evidence, and it's the same bar multi-vendor agents will have to meet next. We joined the MACH Alliance to bring real solutions and experience into agentic architectures both within the U.S and internationally."

His comments reflect a broader debate in the AI market over whether organisations can rely on autonomous or semi-autonomous systems without being able to inspect the chain of reasoning and data behind them. For sectors with strict compliance requirements, explainability is increasingly being treated as a practical requirement rather than a technical preference.

Holly Hall, Managing Director of MACH Alliance, linked Seekr's arrival to that need for oversight as well as interoperability. "We're pleased to welcome Seekr to the Agent Ecosystem as a Collaborator," Hall said. "As enterprises connect agents across vendors, the ability to trace and explain what those agents decide becomes as important as getting them to work together. Seekr's experience in regulated and government environments brings that perspective directly into the working groups shaping our reference architectures."

Seekr says it brings expertise in governance, security and evaluation to the alliance. The company works on explainable AI for organisations that need to audit model decisions, including users deploying large language models, vision language models, edge-based systems and AI agents across different infrastructure environments.

The MACH Alliance's membership includes technology providers and consultancies such as Vercel, Deloitte Digital and Netlify. Its Agent Ecosystem has been positioned as an effort to establish common structures for multi-vendor AI agents before proprietary approaches become entrenched, with traceability now emerging as one of the central design questions.