Operations | Monitoring | ITSM | DevOps | Cloud

Watch the full episode: MCP vs CLI: Does it even make a difference? Live Laugh Logs ep. 3.

When AI could only answer questions, a bad prompt just meant a bad answer. But now AI agents read your documents, browse websites, and actually do things on your behalf. So when someone sneaks a malicious instruction into a file or a webpage, the agent doesn't just say something wrong. It does something wrong!

Highly classified environment. Strict compliance requirements. Data that can't leave the network. But still a real need for the competitive advantage AI delivers. Civo Director of Enterprise Cloud Solutions John Dietz addresses exactly that challenge and how Konstruct makes it possible to run Kubernetes, deploy your own models, and point Claude Code at your own internal private servers instead of public APIs.

AI incident assistant from BigPanda gives L2, L3, and SRE teams instant answers to resolve incidents faster without manual triage or tool-switching. IT teams lose critical minutes during incidents because context is scattered across Slack threads, bridge calls, monitoring tools, and historical tickets. The BigPanda AI Incident Assistant fixes that by surfacing relevant knowledge exactly when and where responders need it. It gives responders evidence-based resolution paths drawn from historical incidents and live system data, without leaving your workflows.

AI Incident Prevention from BigPanda stops change-related outages before they occur by leveraging risk scores, trend analysis, and guided remediation steps. Manual IT changes are still a leading cause of IT outages and disruptions. BigPanda AI Incident Prevention addresses this by automatically scoring change requests against historical data, flagging high-risk changes before they go live, and surfacing the recurring problems that cause service degradation.

On this episode of Masters of Data, we revisit the predictions Adam White, Zoe Hawkins, and David Girvin made at the end of last year, checking our own scorecard halfway through 2026. The hits: agents running amok and deleting databases, MCP becoming the backbone for tracking what agents actually do, growing security gaps around personal data, and a collective rejection of low-quality AI content. The misses: we underestimated how fast companies would cut staff for AI, then quietly start rehiring once the agents couldn't cover the work, and we're still arguing about whether token burn is a cost problem or a coming attack vector.

Claude Code + CircleCI MCP = no more hunting through build logs.