CompanyGraph
Dynatrace, Inc.
DT · NYSE Arca · United States
Installs a monitoring agent on every server and container so its AI can automatically find the root cause of software problems.
Dynatrace installs a small piece of software called OneAgent on every server and container inside an enterprise's IT environment, which captures a continuous stream of execution traces from every application component it touches and feeds them into an AI engine called Davis that automatically draws a live map of how all those components depend on each other. Because Davis builds that map from the trace stream rather than from rules that a human engineer types in, its ability to pinpoint the root cause of an outage is only as good as how completely OneAgent covers the environment — one unmonitored server or container creates a gap where the causal chain breaks. That coverage dependency is also what makes customers hard to move: OneAgent gets woven into the deployment scripts companies use to ship their own software, and the months of learned baselines and dependency relationships Davis has accumulated inside a customer's environment cannot be transferred to a new vendor, who would have to start over from scratch. The main thing that slows Dynatrace down is not engineering capacity but the pace at which large enterprises — navigating security approvals, change management processes, and data-residency rules — can actually get OneAgent installed across thousands of hosts.
How does this company make money?
Companies pay an annual subscription fee based on how many hosts or containers are monitored. On top of that base fee, additional charges apply for specific add-on capabilities — synthetic monitoring, which simulates user actions to test availability; real user monitoring, which tracks what actual users experience; and cloud infrastructure monitoring beyond the core application performance management platform.
What makes this company hard to replace?
OneAgent gets embedded directly into the scripts and templates companies use to deploy their software, so removing it requires reworking those deployment processes. The performance baselines and dependency maps that Davis has learned over months inside a specific environment cannot be exported or reproduced — a new vendor would have to start over from nothing. Custom dashboards and alert rules built on top of Davis AI outputs would also need to be rebuilt entirely with whatever the replacement tool offers.
What limits this company?
Before the AI can reliably trace the cause of any incident, OneAgent must be running on every single host and container in the environment. In large companies, the specific bottleneck is not computing power or the algorithm — it is getting permission to install the agent. Security teams, change approval processes, and old systems that cannot accept new software all slow down how quickly full coverage can be reached across thousands of machines.
What does this company depend on?
The company cannot function without the ability to deploy OneAgent on Windows, Linux, and containerized environments. It also relies on APIs from AWS, Azure, and Google Cloud Platform to monitor cloud infrastructure, and on application runtimes including Java, .NET, Node.js, and Python to instrument code. Kubernetes cluster access is required to monitor containers, and enterprise identity management systems must grant the authentication needed to operate inside customer environments.
Who depends on this company?
DevOps teams at Fortune 500 companies use it to detect incidents and identify root causes automatically during outages; without it, they would have no automated way to know what broke or why. Site reliability engineers would fall back to manually reading through logs and guessing when applications slow down across many interconnected microservices. Teams running cloud migration projects would lose the dependency maps and performance baselines they rely on to move infrastructure safely.
How does this company scale?
Once the AI algorithms and dependency mapping logic are built, adding a new customer environment costs very little — the same software simply runs in a new place. What does not scale cheaply is the human side: specialized customer success engineers who understand complex enterprise systems and can guide a company through getting full OneAgent coverage are hard to automate or replace.
What external forces can significantly affect this company?
GDPR and local data residency laws can force telemetry to stay inside specific country borders, which prevents Davis from analyzing connections that cross those borders. FedRAMP and similar federal cloud security requirements demand lengthy compliance certifications before product updates can be released to government customers. When the broader economy weakens, large enterprises tend to cut costs by consolidating their monitoring tools, and the per-host pricing model becomes a target for those budget reviews.
Where is this company structurally vulnerable?
If a company's security or legal policies block OneAgent from being installed on certain servers — for example, because GDPR rules require data to stay inside a specific country, or because FedRAMP mandates restrict how telemetry moves across regions — gaps appear in the trace stream. When those gaps exist, Davis cannot close its causal chains. The automatic root-cause capability then degrades into the same incomplete, manually-patched process that every competitor already offers.