The Second Coming Of AI [Part I]
By Jeremy BurtonApril 18, 2023
We believe these technologies will have an immediate and profound effect on the adoption of Observe, so much so that we’ve already released our first feature, O11y GPT, which is already in production and is being used by our customers today.
Meet O11y, Powered By GPT
For years, we struggled to find real-world applicability for AI/ML in the realm of Observability — that is, until recently, with the advent of Large Language Models (LLMs) and ChatGPT. We believe these technologies will have an immediate and profound effect on the adoption of Observe, so much so that we’ve already released our first feature, O11y GPT, which is already in production and is being used by our customers today.
Today, LLMs are clearly not at the “train your model on petabytes of machine-generated data and auto-magically find problems” stage. Nor are they the early underpinnings of HAL 9000 from 2001: A Space Odyssey either. However, they are something that can generate surprisingly coherent sentences, code snippets, etc and, with the right training data, can be surprisingly correct.
Better Help, Through O11y GPT
The most obvious and immediate place to use GPT technology was to provide better, more accessible help. Enterprise software has a depth of features, unlike anything you’ll find in the consumer world. 
Sure, you can make simple things easy, but there’s always a point at which the user has to learn — and that means reading documentation, watching a video, or studying an example.
Throughout my career, I’ve always been underwhelmed by the technical documentation of the products I’ve worked on. They were either too dense (written by engineers), too fluffy (written by marketers), or too confusing (written by doc writers).
So what if we could feed an LLM with perhaps dense, but technically correct, materials and have it generate coherent sentences that anyone can understand? That would obviously be a huge benefit to new users and those that only use the system once or twice a week. But it would also be beneficial to the power user who is looking for examples of an OPAL verb they don’t use too often.
The dream of any Product Manager in the world of enterprise software is for a user to log in and be productive — easily accomplish a task or two — within the first 30 minutes of using their product. Today we interrupt the user with wizards, tutorials, guided tours, and videos, but it would be so much better if the user could just tell the product what they wanted to do — and get precise instructions on how to do it.
That’s what O11y GPT does, and it’s available today.
What Next?
Better help is a great start, but that’s all it is — a start. We have an amazing roadmap of GPT-related features that we will roll out in the coming weeks and months. Our next blog will take things to the next level highlighting new features to make scripting in OPAL a breeze — even for novice users.
Ready to revolutionize your observability practice? Click here to unlock the benefits of our latest GPT-related integrations like O11y GPT (and more) for your team!