Functional Agents and Alert Investigations
The May 2026 Aiden release covers Functional Agents, Observability and Alert Integration.
Explore the sections below to see what's new, enhanced, and fixed in this update.
| Feature | Link | |
|---|---|---|
| What's New | Agents | Functional Agents |
| Integrations | Observability and Alert Integration |
What's New
Functional Agents
Click to view
Functional Agents (also referred to as personal agents) are customizable AI entities in Aiden that you can configure with specific expertise and operational guidelines.
Key Features
-
Persona customization - Define specific roles for an agent, such as a Cloud Infra Architect, to handle specialized tasks.
-
Expert integration - When you define an agent, select which expert it is allowed to work with so the agent uses the correct domain knowledge.
-
Custom instructions - Provide unique instructions to each agent to govern how it responds and the standards it should follow.
-
Thought process visibility - When an agent generates output (for example Terraform code), its internal reasoning and thought process are visible, including which agent produced the response.
-
Agent filtering - Filter your list of agents by text to quickly find the right persona for a conversation.
Why It Matters
Specialized personas and transparent reasoning give you more accurate, context-aware assistance, so teams can generate infrastructure code and troubleshoot complex environments according to their own organizational standards.
Observability and Alert Integration
Click to view
Aiden now provides enhanced visibility and automated context-gathering for infrastructure alerts and investigations.
Key Features
-
Full alert payloads - For environments with Grafana or Observe configured, Aiden can display the complete information payload for any captured alert.
-
Investigation detail visibility - The Investigation tab now renders all detailed data related to a specific infrastructure investigation in one view.
-
Contextual conversations - Start a new Aiden conversation that automatically includes the full payload of an investigation for more accurate, data-driven assistance.
-
Chain of thought rendering - The internal reasoning and chain of thought for an investigation are visible in both the dedicated investigation view and the main chat window.
Why It Matters
By surfacing full alert payloads and maintaining the chain of thought throughout the investigation, Aiden gives DevOps teams immediate, deep context. This reduces the time spent manually gathering data from external observability tools during critical incidents.