Mezmo Honored with Open Agentic AI Innovation Award presented by the Agentic AI Foundation

The award recognizes Mezmo’s work on AURA, an Apache 2.0 open-source harness built to address a problem the company encountered firsthand: how to give agents meaningful access to production while keeping their actions within boundaries defined by the operator.

As organizations move agents from experimentation into production, that challenge is becoming increasingly important. The question is no longer simply whether an agent can reason about an original problem, but whether teams can safely delegate meaningful production work without giving up control over how that work is performed.

Read more – Avahi Achieves AWS Security Competency for Expertise in Protecting Customer Environments

Mezmo’s experience operating at petabyte telemetry scale helped expose the limitations of general-purpose agent approaches. As its SRE team experimented with agents for incident response, investigations could overwhelm model context, consume expensive tokens on routine work, require repeated approvals, and create pressure to give agents broader permissions than operators were comfortable granting.

AURA was built around a different model. Probabilistic reasoning happens inside deterministic boundaries defined by the operator.

Teams control which tools and systems an agent can access, where human approval is required, and what evidence must be preserved. Those controls sit outside the model, so the agent cannot expand its own permissions through prompting or reasoning.

For SRE teams, that means an agent can start from an alert, investigate logs and metrics, connect an issue to a deployment or code change, and recommend or take action within those boundaries. The evidence behind its decisions and its execution history remain available so operators can understand what happened, why it happened, and what the agent did next.

Read more – Zerberus Technologies Wins 2026 AI Security and Governance Award  

“We are honored to receive the Open Agentic AI Innovation Award,” said Tucker Callaway, CEO of Mezmo. “We built AURA because making an agent smarter does not make it safe to operate in production. Production requires a system around the model that controls its authority, manages its context, preserves evidence, and gives operators the final say where it matters. We believe that foundation should be open, inspectable, and owned by the teams running it.”

Read more – ACL Digital Launches AgentixCare AI Platform for Healthcare

AURA is open source by design rather than a proprietary agent tied to a single model or service. Organizations can run it in their own environments, inspect and extend its behavior, choose their own models, and define the boundaries within which agent operates.

This creates a path toward increasing autonomy deliberately. Teams can begin with investigation and recommendations, introduce human-gated actions, and expand agent authority as workflows become better understood and trusted.