For the better part of a decade, the observability sector has promised that machine intelligence would eventually take the grunt work out of running complex software estates. Vendors sold dashboards that lit up when something broke, then sold more dashboards to make sense of the first ones. Now one of the biggest names in the field says the technology has genuinely turned a corner, and it is reorganising the way it sells to prove the point.
In an interview with CRN Australia, Dynatrace’s vice-president of worldwide partners argued that “AI has now come of age” inside the observability market, a line that doubles as both a product claim and a channel strategy. The message to the resellers, managed service providers and consultancies that carry Dynatrace into customer accounts is that the machine-learning features baked into the platform are no longer a nice-to-have demo, but the reason the sale happens at all.
The context: from dashboards to decisions
Observability is the discipline of understanding what is happening inside modern software from the outside, by collecting the logs, metrics and traces that applications throw off as they run. As organisations moved to cloud platforms, microservices and containers, the volume of that telemetry exploded well past the point where a human operations team could reasonably eyeball it. That is the gap the industry has spent years trying to fill with automation, often marketed under the clumsy label AIOps.
Dynatrace has staked much of its reputation on an engine it calls Davis, which it uses to spot anomalies, work out the root cause of an incident and, increasingly, suggest or trigger a fix. The company’s pitch has recently folded in generative AI as well, giving engineers a way to interrogate their systems in plain English rather than by writing queries. The argument that AI has come of age is really an argument that these two strands, the predictive machine learning that has been maturing for years and the newer generative layer, have finally converged into something a customer will pay a premium for.
The news: a channel built around AI
What makes the comments more than marketing is the structural change sitting behind them. Dynatrace runs the bulk of its business through partners rather than selling everything direct, so a claim that AI is now the centre of gravity has to be matched by partners who can actually deploy and explain it. The executive’s framing suggests the vendor wants its channel to move up the value chain, away from simply reselling licences and towards designing the automated workflows, guardrails and integrations that let AI features do useful work in a customer’s environment.
That is a meaningful shift for the smaller Australian firms that make up a good slice of any global vendor’s local ecosystem. Reselling software is a volume game with thin margins. Building and maintaining the automation around it is a services game with much better economics, provided the partner has the skills to pull it off. The come-of-age narrative is, in part, an invitation to partners to invest in exactly those skills.
Two ways to read it
The optimistic reading is straightforward. Enterprise IT teams are drowning in alerts and short on staff, and anything that reliably reduces the noise and shortens the time to resolve an incident has obvious commercial value. If Dynatrace and its rivals have genuinely crossed from novelty to dependable tooling, then the operations teams keeping banks, telcos and government services online stand to benefit, and the partners who implement it well should do nicely.
The sceptical reading is worth holding onto too. “AI has come of age” is a claim vendors across the observability field are making almost word for word, including Datadog, New Relic, Splunk and Grafana, each with its own branded intelligence layer. For customers, the harder questions are not whether the AI is clever but whether it is trustworthy: does it surface the right root cause often enough to be relied on, how much of its reasoning can an engineer inspect, and what happens when an automated remediation acts on a false positive during a live incident. Coming of age implies a level of accountability that the marketing does not always spell out.
There is also a cost dimension that Australian buyers have grown wary of. Observability platforms bill on the volume of data ingested, and as AI features encourage teams to collect and analyse ever more telemetry, the invoices can climb fast. FluentSea has previously reported on the phenomenon of AI “bill shock”, and observability is one of the clearest places it shows up.
The Australian stakes
For Australia, the timing matters. Local enterprises and government agencies are pushing workloads into the cloud at pace, often across hyperscale regions in Sydney and Melbourne, and they are doing it while carrying real skills shortages in site reliability engineering and cloud operations. Tooling that can genuinely absorb some of that operational load is attractive precisely because the people to do the work by hand are scarce and expensive.
The flip side is regulatory. Sectors such as banking, health and critical infrastructure operate under the SOCI regime and APRA’s CPS 230 operational-resilience standard, which demand that organisations understand and can account for how their critical systems behave. Handing incident response to an automated engine sits awkwardly with rules that expect a clear chain of human accountability. Australian buyers will want to know they can see inside the AI’s decisions, not just accept its outputs, before they let it act on production systems. That is likely to be the real test of whether the technology has come of age in the local market, rather than any vendor’s say-so.
For the channel, the opportunity is concrete. Australian managed service providers that can wrap observability AI in the governance, documentation and human oversight that regulated customers require will be selling something more defensible than a licence key. Those that cannot will find themselves competing on price in a shrinking margin pool.
What’s next
Expect the come-of-age language to keep spreading across the sector through the rest of the year, alongside a steady drip of new agentic features that promise to move from suggesting fixes to executing them. The interesting signal to watch in Australia will not be the announcements but the procurement conversations: how customers price the data ingestion, how they audit the automation, and whether partners can demonstrate the skills to make any of it stick. Maturity, in the end, is measured by customers, not by the vendors declaring it.
Sources: CRN Australia.
