Operations | Monitoring | ITSM | DevOps | Cloud

Single agents are a useful starting point for SRE workflows. They are not where the architecture should end. The first version is simple enough: connect an LLM to a few tools, give it a system prompt, and point it at your infrastructure. It can summarize an alert, pull logs, answer questions, and draft a useful next step. Then the workflow gets real. You add GitHub for runbooks, Kubernetes for cluster state, PagerDuty for incident context, Prometheus for metrics, and Mezmo for telemetry.

Last week, we introduced Agent Timeline, a powerful new observability experience purpose-built for debugging AI agent workflows in production. Agent Timeline uniquely connects AI-layer visibility to full-stack observability by organizing telemetry around an agentic conversation. A conversation contains one or more agent executions, each of which may contain LLM calls, tool invocations, handoffs, retries, human escalations, and downstream system calls.

The average person would need 180 million years to read everything published online in a single day. For organizations trying to track what people say about their brand, manual monitoring stopped being viable somewhere around 2015. AI-powered media monitoring tools now process this impossible volume automatically, detecting brand mentions, analyzing sentiment, and flagging potential crises before they spiral.

Your dashboards are green. Your thresholds are calm. Then a cascade failure starts and you don't know until users flood your status page. Traditional monitoring is reactive by design. Anomaly detection in observability changes that equation entirely.

When a PostgreSQL query runs slowly, engineers often start with EXPLAIN ANALYZE. The output is a tree of plan nodes, each one describing a step the database took to execute it. A query with several joins and a subquery can produce 20 or more nodes. But the plan gives no visual indication of which node corresponds to each clause in the SQL text. Diagnosing the problem means viewing the plan in one window and the query in another, manually tracing connections between them.

Metric exploration often begins with a simple question, but answering that question can require deep familiarity with metric names, tag structures, and query syntax. Experienced users spend time refining queries through trial and error, and newer users struggle to get started. As a result, teams face delays in troubleshooting and analysis. Valuable observability data, including metrics that are difficult to discover and query, also goes underused.

“Autonomous” doesn’t mean uncontrolled. AI-driven IT in life sciences needs guardrails based on GxP, risk, and validation. Low-risk systems can automate; high-risk systems need human approval and documentation.

Today we're rolling out a major upgrade to API Keys in Bindplane. You can now create up to 25 API keys per project, give each one a description, set an expiration date, and delete keys you no longer need. Under the hood, every key is now hashed with Argon2, the modern standard for credential storage. If you've been working around the old single-key limit by sharing one key across CI jobs, scripts, and teammates, this release is for you.

One of the things we love most about building an open source platform is seeing what the community creates with it, and independent developer Anton Havekes recently built something we just had to share. Anton put together Influx Dashboard, a native iOS app that connects to your InfluxDB instance and brings your time series data straight to your phone. We’re genuinely thrilled to see this kind of work come out of the community.