The name Lew Cirne is actually an anagram of the title New Relic, showcasing a clever wordplay. However, it is suggested that a more fitting name aligned with the company’s mission as an all-in-one observability platform could have been “Crew Line,” utilizing the same letters while maintaining the anagrammatic connection.
New Relic, a prominent figure in Application Performance Management (APM), aims to unify all software engineering team members (referred to as the crew) to streamline the creation, deployment, and maintenance of applications throughout the software supply chain.
Significance of AI APM
This month, New Relic introduced New Relic AI Monitoring, an APM service tailored for AI-driven software. The integration of APM in AI is crucial due to the evolving landscape of AI technologies, differing from traditional APM practices. Manav Khurana emphasized the importance of observability in AI development and operation, providing professionals with the tools to comprehend the intricacies of AI and develop effective programs securely.
The key focus lies in understanding and managing the data flow within AI applications, particularly the Large Language Models (LLMs), to evaluate their value, robustness, security, and reliability accurately.
New Relic’s approach involves merging AI observability with AI tracking to grant software engineers comprehensive oversight of AI operations, facilitating troubleshooting and enhancements. With support for over 50 interfaces across the AI spectrum, including popular LLMs like OpenAI GPT-4, the company’s AI monitoring technology boasts the capability to monitor diverse AI ecosystems effectively.
Distinction in AI Monitoring Techniques
AI-powered technologies introduce complexities as elements such as LLMs and variable data stores often pose challenges for engineers, generating vast amounts of monitoring data that require meticulous analysis and monitoring to mitigate potential security risks.
Artificial monitoring enables engineers to optimize LLM interactions, addressing efficiency, cost, security, and quality concerns like bias, hallucinations, and toxicity. By offering complete visibility into all AI stack components, services, and infrastructure, New Relic empowers engineers to ensure compliance with AI regulations effectively.
The integration of various solutions like LangChain for instrumentation, LLMs such as MosaicML and HuggingFace, machine learning frameworks like TensorFlow and Pytorch, and AI models like Amazon SageMaker and AzureML, among others, enhances the monitoring process. Additionally, visibility across the entire AI application stack, encompassing AI metrics and APM golden signals, is provided without additional instrumentation requirements.
Collaboration with Rock Amazon
New Relic’s collaboration with Rock Amazon, an AWS-managed service that facilitates access to foundation models from leading AI companies, enhances the AI Monitoring product’s capabilities. This integration allows AWS customers to leverage New Relic for comprehensive insights and optimization across their AI applications, ensuring enhanced performance, quality, and cost-efficiency.
In essence, leveraging AI for application development and performance management, including the monitoring of AI applications, is pivotal in ensuring operational efficiency and adherence to quality standards. This holistic approach underscores the essence of Application Performance Management in the realm of AI technologies.
