Dynatrace expands its platform with AI Observability, a new solution that enables organizations to monitor and control every layer of AI-driven applications. This addition gives Dynatrace a comprehensive offering for observability and security tailored to large language models (LLMs) and generative AI applications. The enhancement helps companies adopt generative AI in a secure, cost-effective way while improving performance, innovation, productivity, and revenue.
“Generative AI holds enormous potential, but we must ensure AI investments succeed and avoid unpredictable behaviors, hallucinations, and poor user experiences,” says Johan Dahlin, Head of Nordics at Dynatrace.
“This further reinforces our leadership in application security and observability. Legacy tools have produced fragmented ways of managing digital channels and applications. With Dynatrace AI Observability, teams no longer have to contend with fragmented data, processes, and limited insights—everything is unified on a single platform. That hasn’t been possible before,” Dahlin continues.
Dynatrace AI Observability leverages Dynatrace’s Davis AI and other core technologies to deliver precise, end-to-end visibility into AI-driven applications. This unified approach improves user experience while automatically detecting performance degradations, security risks, and root causes.
The solution also simplifies compliance with privacy, security, and regulatory standards by tracing the origin of data produced or consumed by applications. Additionally, it helps forecast and control costs by tracking consumption of tokens—the unit generative AI models use to process queries and which drive usage costs.
Cost concerns for generative AI services
According to Gartner, by 2028 more than 50 percent of cloud computing resources will be devoted to AI workloads, compared with under 10 percent in 2023.
This growth reflects rising interest in generative AI to boost efficiency, productivity, automation, and innovation. Still, many companies and organizations worry about the costs tied to generative AI services. These services can be substantially more expensive than traditional cloud offerings and often hard to predict because they’re driven by consumption of generative AI tokens and by applications that are not yet in production. At the same time, regulators worldwide are introducing rules to ensure AI is used responsibly, ethically, and in accordance with applicable laws.
“Generative AI enables innovative solutions that increase productivity, profitability, and competitiveness. But it also creates new challenges around security, transparency, reliability, user experience, and cost control. Organizations therefore need complete visibility across all their generative AI deployments, and we’re expanding our observability and AI capabilities to meet that need,” says Bernd Greifeneder, CTO at Dynatrace.
“AI-driven transformation is a key driver of productivity and competitiveness worldwide. Microsoft Azure OpenAI Service is a robust enterprise-grade generative AI offering, and we are pleased to work with Dynatrace—an AI-driven observability leader—to help our mutual customers build generative AI services using Azure OpenAI Service and Dynatrace AI Observability. Together, these solutions help ensure AI services meet the highest standards for security, reliability, and performance while enabling teams to manage costs,” says Ali Dalloul, VP of AI at Microsoft.
Dynatrace AI Observability covers the entire AI stack, from infrastructure such as NVIDIA GPUs to foundational models like GPT-4, semantic caches, and vector databases such as Weaviate. It also supports major platforms for building, training, and serving AI models, including Microsoft Azure OpenAI Service, Amazon SageMaker, and Google AI Platform.
Dynatrace AI Observability is available to all Dynatrace customers. For the current list of supported technologies and integrations, visit the Dynatrace Hub on Dynatrace’s website. For additional details about the solution and its capabilities, see the Dynatrace blog post about Dynatrace AI Observability.