Operations | Monitoring | ITSM | DevOps | Cloud

Findings from Digitate’s recent survey conducted with Sapio Research highlight a persistent challenge: AI is often implemented to reduce human workload and operational costs, yet these very factors continue to limit its broader adoption.

If you’ve ever opened the alerts tab during a busy period, you know the problem. There are alerts you’ve already looked at, alerts someone on your team is handling, and alerts that fired on a known issue that’s being worked on. They all sit together in the same list alongside the new ones you haven’t seen yet.

Runbooks are rarely missing because teams don't value them. They're usually missing because incident response, follow-up, and platform work compete for the same limited time. By the time an issue is resolved, the knowledge is fresh, but the window to document it is already closing. That gap creates familiar failure modes: over-reliance on senior engineers, slower handoffs, and less confidence for whoever is on call next.

On April 1st, I joined Akshay Utture from Augment Code for a webinar on how AI agents use production feedback to improve code.

Autonomous IT is the result of a long operational evolution, from static monitoring and rule-based automation to AIOps and now to systems that can increasingly diagnose, prioritize, and act within defined guardrails. Autonomous IT gets talked about like it appeared out of nowhere. As if someone flipped a switch and suddenly systems started managing themselves. The reality is far less dramatic and far more instructive. What we’re seeing today is the result of decades of incremental progress.

Effective network monitoring requires path validation, not only device polling. Traditional Network Monitoring System (NMS) tools were built for static networks, not today’s hybrid reality. You poll devices, check interface counters, and still struggle to explain why users complain about latency. Traffic moves across SD-WAN architectures, cloud routing layers, and public internet paths that device metrics never capture.

Fragmentation limits AI automation because context is split across systems, forcing humans to bridge the gap. Most IT environments are fragmented by design. Observability data lives in one set of systems, investigation happens in another, and execution sits behind separate tools with their own ownership and controls. During an incident, context does not move with the work.