New Relic launches Autopilot SRE agent and "Ground Truth" for customer agents

What changed: New Relic announced New Relic Autopilot, an out-of-the-box SRE (site reliability engineering) agent that triages incidents and proposes remediations, and New Relic Ground Truth, which gives customers’ own agents direct, agent-optimized access to New Relic observability data.

Why it matters: Engineering teams can either let New Relic run a continuously operating SRE agent for incident handling or connect their existing AI agents (Copilot-like or homegrown) to richer, purpose-built telemetry so agents make faster, better-grounded actions. That reduces manual incident triage and improves how agents interpret system state.

Try / watch: If you run services, evaluate a pilot that connects a low-risk alert type (e.g., failed cron jobs, noncritical latency spikes) to Autopilot or Ground Truth to measure false positives, time-to-remediation, and cost of agent queries before widening usage.

Zafin debuts AIOS: an orchestration and governance platform for agentic work

What changed: Zafin released AIOS, an end-to-end platform that registers, orchestrates, and enforces governance for an institution’s agents, third-party agents, models and tools across full business workflows, with built-in proof-of-work and cost controls. The announcement frames AIOS as targeted at regulated firms that need auditable execution paths.

Why it matters: For regulated businesses (banking, insurance, healthcare) the immediate problem isn't agent capability but traceability and policy enforcement; AIOS promises a way to run agentic automation while keeping human approvals, evidence, and cost controls in place. That can make multi-step, cross-system agent workflows usable for higher-risk operations.

Try / watch: Regulated operators should map one high-value, audit-sensitive workflow (for example, a pricing change or customer data update) and run it through AIOS’ proof-of-work to validate that audit records, human checkpoints, and cost accounting match compliance needs.

Expel extends Ruxie (AI SOC manager) with agentic capabilities across the threat lifecycle

What changed: Expel expanded Ruxie, its AI SOC manager, to apply agentic workflows across enrichment, detection, triage, investigation, automated response actions, rule engineering, and reporting in production for customers. The company says these capabilities are live in customer MDR deployments.

Why it matters: Security teams face AI-powered attacks that move at machine speed; packing coverage into agentic workflows—especially automated enrichment and response—lets defenders compress detection-to-remediation timelines and free analysts to handle high‑uncertainty edge cases. But it also raises the need for clear human‑in‑the‑loop gates and rollback paths.

Try / watch: SOC leaders should insist on measurable safe‑guards: test Ruxie agentic responses in staged environments, verify rollback procedures, and monitor whether automated triage reduces analyst workload without increasing missed incidents or false closures.

Solink announces general availability of AI Agents for video‑centric, outcome-driven automation

What changed: Solink made its Solink AI Agents generally available, positioning prebuilt and custom agents that reason across video and operational data to drive actions like loss prevention, overnight monitoring, and visual compliance at scale. The launch was revealed at Solink’s Agentic AI Summit.

Why it matters: Physical‑world operations (retail, hospitality, property management) can move from dashboards to automated, outcome-focused agents that alert the right person with validated context—reducing shrink, improving safety, and cutting manual review time. Buyers should validate precision on their stores/cameras before trusting autonomous actions.

Try / watch: Operators should pilot a single prebuilt agent (e.g., Loss Prevention) at a few locations, measure true positive rate and operational response time, and confirm integration with existing workflows (alarms, incident tickets, guards) before scaling.