Observe Raises $50M and its Olly GPT is a Generative AI Assistant for Infrastructure Monitoring
Generative AI has broad application
Observe has announced a new $50 million funding round in convertible debt led by Sutter Hill Ventures. This brings Observe’s total funding to $164.5 million. TechCrunch reported that the debt will convert to equity next year at the time of the company’s Series B funding round.
The financing will be used to grow Observe’s sales and R&D teams, according to CEO Jeremy Burton, as the company looks to expand its headcount from 150 employees to 250 by the end of 2024.
Burton says that the decision to opt for debt funding was in the interest of deferring dilution.
“We expect to raise the Series B early next year, at which time the debt will convert to equity,” Burton told TechCrunch in an email interview. “We’ve run the company off of debt financing for the past three years.”
Generative Observability
Logs. There is a lot of useful data contained in logs for operational systems. However, it is a classic problem of identifying signals in noise. It’s not that organizations don’t have access to data. In 2023, application and infrastructure telemetry data are plentiful. The problem is too much data spread across too many systems. “Pains of glass become pains in the a**” is a phrase you might hear from IT professionals who spend their days mired in monitoring dashboards.
Earlier this year, Observe introduced the OPAL Copilot to enable users to generate code that executes commands based on a fine-tuned GPT model. The company also created O11y GPT [pronounced Ollie GPT], a generative AI powered assistant for users. O11y GPT started out as a chatbot for help questions. More recently, it has added features that make analysis and taking action easier for IT operations.
O11y GPT [pronounced Ollie GPT] is Observe’s generative AI assistant to help users troubleshoot incidents faster. Observe users can ask questions in plain English using a familiar chatbot interface to fetch data extract fields and make sense of error messages and much more. O11y GPT can even act as an assistant for on-call engineers to remediate an incident via Slack. O11y GPT answers users’ questions about how to perform common tasks such as finding relevant data sets or what, the configuration to use in order to ingest Prometheus metrics…
With Hubble, O11y GPT provides users with more data insights for example highlighting an error message in a log now explains that error message regardless of source…
In addition, new logs often arrive unstructured, so users may have to write complex regex statements to parse them before they can even perform analytics. This can be a nerve-wracking experience during an incident when time is precious. O11y GPT not only generates the regex but also names the resulting columns appropriately such as URL or status code.
The presentation continues with additional use cases and features [start at 10:50], but this is sufficient to make the point. When you first saw GPT-3, or even ChatGPT, did you think it would change cloud or application monitoring/observability?
The applications of generative AI are far broader than is often assumed. Even use cases with a lot of structured data, such as monitoring, also have a significant amount of unstructured data in code or language formats. Plus, the ability to add an assistant that can respond to natural language requests to just about any software application has its merits.
What This is Not
It is worth clarifying that Observe is not Observe AI. The former focuses on monitoring cloud infrastructure and applications. The latter is a conversational intelligence solution for contact centers. Today, both use large language models (LLMs) but for different use cases.
What’s Next
It will be interesting to see how adoption progresses among Observe’s customers. You might think that power users are the least likely to want the help of an AI assistant. However, we have seen in other market segments that the most proficient with an application, coding language, and expertise domain, tend to get even more value than their peers when generative AI is added to the mix.
Observe noted in its annual industry report that 61% of IT teams surveyed had “experienced reduced budgets, and 55% have had to reduce headcounts.” Generative AI is coming along at a time when companies need to raise productivity just to maintain the status quo. In many ways, generative AI productivity gains are viewed as a revolution of necessity for enterprises.
Let me know what you think in the comments below.