Dynatrace

What it is

Dynatrace is a commercial, full-stack observability and application security platform founded in Austria in 2005 (as a spin-off of an earlier APM tool) and now headquartered in Massachusetts. Its distinguishing technology is OneAgent, a single agent that auto-discovers processes on a host and auto-instruments applications without manual configuration, combined with PurePath, its distributed tracing technology that stitches every request into an end-to-end trace enriched with code-level detail, topology, logs, and user-experience data. Traces, metrics, logs, and topology all land in one unified data model (Grail, Dynatrace’s proprietary data lakehouse) rather than in separate siloed stores.

The platform’s best-known feature is Davis, Dynatrace’s AI engine. Davis performs deterministic, causal root-cause analysis (not just anomaly correlation) by walking the dependency graph built from OneAgent’s topology data, and Dynatrace has been investing in this causal approach since roughly 2018 — longer than most competitors have marketed “AI observability.” More recently Dynatrace has positioned Davis around agentic workflows: autonomous “domain agents” that investigate and, in some cases, propose or trigger remediation.

Why teams choose it

• Zero-configuration automatic tracing. OneAgent instruments the full stack — application code, containers, infrastructure, and increasingly OpenTelemetry-emitted signals — without hand-placed instrumentation, which matters most in large, heterogeneous, frequently-changing environments.
• Deterministic root-cause analysis. Davis’s causal engine, built on a real topology model rather than statistical correlation alone, tends to produce fewer false leads during incidents than tools that surface many “possibly related” anomalies.
• Genuinely unified data model. Metrics, traces, logs, security events, and business data live in one backend (Grail) that can be queried together, avoiding the cross-tool correlation work that fragmented toolchains require.
• Cost and complexity. Dynatrace is consistently one of the more expensive platforms per host/GiB, and its breadth (APM, infrastructure, logs, security, business analytics, session replay) can be more platform than teams that only need tracing actually want; smaller or simpler environments may find OneAgent’s footprint and Dynatrace’s learning curve heavier than necessary.

Pricing model

Dynatrace uses consumption-based pricing under its Dynatrace Platform Subscription (DPS): customers commit to an annual spend level and draw it down against a published rate card covering distinct usage dimensions (host hours, GiB of logs ingested, session counts, and similar units per capability). There is no perpetual free tier for production use, though a trial and a bounded free tier exist for evaluation. Overage beyond a commitment is billed at standard on-demand rates rather than penalty pricing.

Alternatives

The closest commercial full-stack competitors are Datadog and New Relic, both of which similarly unify tracing, metrics, and logs with their own AI/anomaly features, generally at a lower per-unit cost but with less deterministic root-cause tooling. For pure distributed tracing without the full commercial platform, open-source options like Jaeger or Grafana Tempo paired with OpenTelemetry instrumentation are common lower-cost alternatives, at the price of doing your own root-cause correlation work.