New Relic AI (NRAI) is designed to help every member of your team improve observability, not just engineers. We combined OpenAI's large language models (LLMs) with our unified telemetry data platform to create an AI assistant that deeply understands how both your system and New Relic function. New Relic AI helps you demystify the complexity of your stack, solve service outages, and improve your efficiency simply by letting you ask questions in plain language.
Ways to interact with New Relic AI
We help you get to the heart of your system by offering quick and easy ways to ask questions: AI chat, and the NRQL console.
Ask AI questions
You can work with our AI any time you want by using our interactive chat experience. All you have to do is click the Ask AI button available on the top right of most New Relic platform pages.
New Relic AI provides a variety of capabilities to help you:
- Transform natural language into queries (NRQL/Nerdgraph)
- Detect anomalies
- Analyze error logs
- Understand stack traces
- Leverage our docs to understand New Relic platform concepts
Our AI has been trained to communicate with engineers using common questions and plain language. Let's say you run an online store, and you start receiving alerts that response times for your Checkout-service are lagging. If you don't have time to dig into the issue, you can just ask New Relic AI "What's on fire?" or "What should I handle first today?" New Relic AI can tell you what's broken in your system and help you explore data with charts and tables.
Sample questions
If you're struggling with where to begin, here are some prompts.
Our AI chat experience can help get you started using New Relic. If you need help organizing your data and setting up alerts, NRAI can act as a guide throughout your onboarding process. Ask our AI questions like:
- How can I use Terraform to create dashboards?
- How can I improve my alerts setup?
- How can I instrument AWS?
If you want to get to the root cause of latency issues for your app, here are some good questions to ask:
- What are some recent incidents in my system?
- Which infrastructure hosts are showing signs of degradation over the past hour?
Our AI can also be your partner in observability maturity. If you're responsible for ensuring that your app can handle spikes for a big e-commerce event like Black Friday, it's important to understand the state of your system so you can improve it ahead of time. Ask our AI:
- How can I improve the performance of my infrastructure?
- Were unique user sessions up or down last week compared to the week before?
Get help creating NRQL queries
NRQL is a powerful tool you can use to gain insights into systems, but when you're in a hurry you might not have time to create the right query to get the data you need. If you're struggling to get a query to work, ask our AI to help you. You can ask questions like, "How is this service performing?" or "How has this service been performing over the past 24 hours?"
To enable NRAI in the NRQL console:
- Go to the user menu at the bottom left of the New Relic UI.
- Click the Features tab.
- Scroll down to the New Relic AI section and click Enable.
Get help analyzing error logs
Error log lines can be difficult to parse. New Relic AI helps you learn more about errors, query for related information, and ask about next steps.
To ask our AI about an error in your logs Go to one.newrelic.com > All capabilities > Logs > All logs > and search for errors.
To chat with New Relic AI about log errors:
- Go to one.newrelic.com > All capabilities > Logs.
- Click All logs.
- Search for
errors. - Click on any resulting log lines and look for the Explain this error button at the top of the log details panel. Our AI will provide a summary of the error.
You can take your investigation further by telling our AI about your stack and asking follow up questions. For example, you can ask:
- How often is this error happening? Meaning, what's the error rate? Is it 1% or 50%?
- What hosts are throwing these errors? I would like to know is it across all of my hosts or is it just affecting one or two of them?
Analyze stack traces in errors inbox
Using our AI to analyze stack traces will reduce context switching and help speed up triage and analysis.
To chat with AI in errors inbox, go to one.newrelic.com > All capabilities > Errors inbox > and select a workload.
To investigate stack traces in errors inbox with New Relic AI:
- Go to one.newrelic.com > All capabilities > Errors inbox.
- Select a workload.
- Look for the Ask AI button at the top of the log details panel. Our AI will provide a summary of the error.
You can take your investigation further by telling our AI about your stack and asking follow up questions. For example, you can say:
Hi, I have an Azure environment and I am having issues with Redis being periodically unavailable. Can you tell me more about this error? How can I diagnose this issue?Secure data
Data security and privacy are primary concerns for anyone building and maintaining software. We want you to know that anything you ask our AI is safe. NRAI is built with the same security controls and standards that safeguard New Relic customer data. We will never share your data with other accounts or customers. We work with vendors that meet our privacy and security standards, and we carefully assess prospective vendors in these critical areas before engaging their services.
Our AI uses a variety of inputs to improve its responses and always provides the best answer. For example, it uses inputs like our product documentation, a large library of example queries, and your telemetry data. But even though our AI scans your data as needed to provide context and help you build charts and understand response times, that data is never stored by New Relic.
Always learning
New Relic AI is becoming more accurate and responsive the more it's used. We want to build the AI experience you want. When chatting with our AI, please give its response a thumbs up or a thumbs down. If you notice something that isn't right or that could be improved, let us know. Go to the user menu, click Help, then click Give us feedback.
Also, New Relic AI can recognize multiple languages and respond in them. If you have use cases for multi-lingual support, please test them with our AI and give us feedback.
New Relic AI responses are AI-generated suggestions. While efforts are made to keep information and responses up to date, suggestions may not always be accurate, complete, or error-free. We would like to encourage you to have a human verify any information provided by our AI using other sources before any action is taken. AI's suggestions are provided without any warranties or guarantees of accuracy.