Trusted AI for 10 years
Dynatrace Davis® AI has been trusted by Dynatrace customers, including in highly regulated sectors, for almost a decade to assist in their decisions with root cause analysis and forecasting.
The industry’s first hypermodal AI
Dynatrace Davis uniquely combines predictive AI, causal AI, and generative AI for observability, security, and business use cases.
Responsible AI in Dynatrace
You can start to use Davis AI out-of-the-box, and explore its full benefits with Dynatrace’s mature approach to AI responsibility by design and default, including:
Control and transparency
Davis AI focuses primarily on systems and empowers users with observations, which they can review to make well-informed decisions. In particular, Dynatrace users benefit from extensive transparency into Davis AI’s analysis, and into the systems Davis AI is evaluating. At all times, customers are in control of the data that is ingested by the Dynatrace platform and considered by Davis AI, Davis AI configuration, as well as any decisions they make based on Davis AI observations.
Security and reliability
As a core function of the Dynatrace platform, Davis AI benefits from our extensive secure development lifecycle and security controls, our regular independent security audits (FedRAMP, ISO 27001, SOC 2), and our regular independent penetration testing (annual external penetration testing, continual Hackerone external penetration testing, and external red team penetration testing on our secure development lifecycle). Davis AI has been extensively tested with each release for almost a decade.
Our responsible AI principles
Responsible AI aims to maximize the benefits to persons from AI, while reducing potential risks.
Human benefit with human control, transparency, and fairness
With digital transformation people rely on software working perfectly to reach their goals and meet their needs. Davis AI focuses primarily on the analysis of systems, not people, to identify system anomalies, perform system root cause analysis and offer system scaling forecasting to help software work perfectly for people.
Humans control each phase of the Davis AI lifecycle, with extensive audience-focused transparency to act effectively on that control. Customer users can use this control and transparency to protect privacy and make decisions about systems that avoid bias and promote fairness.
Causal AI and Predictive AI are trained only on the data of the specific customer using Davis AI, and only assesses the data of that customer. Customers have full control and transparency over the data Davis AI considers, including by selecting the systems that Davis AI assesses, by selecting the data Davis AI assesses in those systems, and by excluding specific data from assessment through the use of rich Dynatrace platform masking features at capture and at ingest.
Customers have full control of all decisions. Davis AI offers observations about systems that customer users can independently assess and verify with the benefit of extensive transparency into Davis AI’s analysis, to support explainability and interpretation. Customer users can perform the assessment and verification with the benefit of the further extensive transparency that Dynatrace observability offers into the same systems, and independent reviews by customers of the same systems they have access to directly.
Accuracy, reliability and robustness
Davis AI undergoes unit and regression tests before each release to ensure repeatability and deterministic results, and to avoid regressions when enhancing the algorithms. Davis AI uses the exact context provided by the OneAgent® and Smartscape®, which are also tested thoroughly before each release. Davis AI has been a core function of the Dynatrace platform since years and benefits from the robustness the Dynatrace platform provides customers.
Security
As a core function of the Dynatrace platform, Davis AI benefits from our extensive secure development lifecycle and security controls, our regular independent security audits (FedRAMP, StateRAMP, ISO 27001, SOC 2), and our regular independent penetration testing (annual external penetration testing, continual Hacker One external penetration testing, and external red team penetration testing on our secure development lifecycle).
Accountability
Accountability is clearly identified at each phase of the Davis AI lifecycle. Customers are responsible for the decisions they make based on the observations provided by Davis AI as they have full control of: the data Davis AI trains on, the data assessed by Davis AI, all decisions they make when they consider the observations provided by Davis AI and extensive transparency into the assessment of Davis AI, and the systems Davis AI assesses.