In today's technology-driven world, we often find ourselves in a dual position as both users and scrutineers of the very technology we rely on. This is especially true when it comes to the intersection of Internet of Things (IoT) devices, cameras, sensors, and tracking tools. While we appreciate the positive impacts they bring, we also want to ensure our personal privacy and anonymity are protected.
A similar duality exists in the realm of Artificial Intelligence (AI). We recognize that AI is constantly monitoring our activities, choices, and actions, both online and in the real world, for the greater good. However, we also need to closely examine AI systems to understand their origins, behavior, and decision-making processes. This practice is known as AI observability.
Given the experimental nature of AI deployment in many enterprises, the field of AI observability is still evolving. Nonetheless, technology vendors are already developing analytics tools and comprehensive observability services to address this critical need. Among these vendors is Dynatrace, a unified observability and security company with an advanced analytics and automation platform.
When Dynatrace claims to provide a unified and holistic view, it carries significant value in this context. Their observability spans across data, code, IT connections (including APIs and containerized software components), the web, the cloud, and even generative AI and Large Language Models (LLMs). With the rise of generative AI, Dynatrace has augmented its platform to incorporate observability specifically for this domain.
According to Dynatrace's CTO, Bernd Greifeneder, generative AI is the new frontier of digital transformation. It empowers organizations to develop innovative solutions that enhance productivity, profitability, and competitiveness. However, this technology also brings new challenges related to security, transparency, reliability, user experience, and cost management. To help address these challenges, Dynatrace has extended its observability and AI capabilities to provide unparalleled insights into generative AI-driven applications, enabling organizations to embrace AI confidently and securely.
Dynatrace's AI Observability is designed to cover the entire AI stack. This encompasses not only the infrastructure, such as hardware components like Nvidia GPUs but also foundational models like GPT4 and vector databases like Weaviate. It also supports popular AI platforms, including Microsoft Azure OpenAI Service, Amazon SageMaker, and Google AI Platform. By leveraging core technologies like Davis AI, Dynatrace provides a precise and complete view of AI-powered applications. This allows organizations to optimize user experiences, identify performance bottlenecks, and automatically determine root causes.
The question that naturally arises is how confident can software engineers be when developing AI applications, and how confident can users be when integrating these smart apps into their daily lives? According to Steve Tack, SVP of product management at Dynatrace, while AI is still in its early stages of deployment, it is continually evolving. Dynatrace AI Observability is built to help deploy performant and secure AI applications. Given that AI functions are typically part of larger services, understanding the dynamic nature of AI technology is essential.
AI observability is also about tracing the provenance of AI output, ensuring compliance with privacy, security, and governance standards. Dynatrace AI Observability, powered by Davis AI, aids organizations in achieving this compliance by precisely tracking the origins of their application's output. Additionally, it assists in forecasting and controlling costs by monitoring the consumption of tokens, which are the basic units used by generative AI models to process queries.
To highlight a recent partnership, Ali Dalloul, VP of AI at Microsoft, emphasizes how Azure OpenAI Service aligns with Dynatrace AI Observability to offer shared customers comprehensive insights. This collaboration ensures that these services meet the highest standards of security, reliability, and performance, while providing cost control for managing teams.
According to Gartner, by 2028, over 50% of cloud compute resources will be devoted to AI workloads, up from less than 10% in 2023. This suggests that many organizations are concerned about the costs associated with generative AI-powered services, which can be significantly more expensive than traditional cloud services. Additionally, forecasting costs can be challenging since they depend on the consumption of generative AI tokens by applications not yet in production.
With governments worldwide establishing regulations to govern the responsible and ethical use of AI technologies, the need for observing and monitoring AI components has never been more critical. Dynatrace's AI Observability empowers organizations to navigate this evolving landscape with confidence and ensures compliance with applicable laws. By adopting AI observability practices, we can instill trust in AI applications while enjoying their benefits both at home and in the workplace.
Our relationship with technology has transformed since the pre-millennial era. We now have technology at our fingertips, granting us unprecedented access and convenience. However, this convenience comes with increased awareness and scrutiny of the platforms and tools we use. Today, we analyze, question, and observe the technology that shapes our lives. Fortunately, there are monitoring and observability tools, like those offered by Dynatrace, that provide a deeper understanding of AI systems and technologies. Through AI observability, we can unlock the full potential of AI while ensuring its responsible and secure deployment.
