Blogs 
Read the original article here.
The landscape of observability tools has undergone another transformation. As the market for solutions aimed at ensuring the reliability of tech systems has expanded over the years, the primary focus has gradually shifted from “monitor everything” to “manage complexity and costs.” Additionally, the swift influx and adoption of AI agents in enterprises have introduced an entirely new category of workloads that demands attention.
InsightFinder AI, a startup founded on 15 years of academic inquiry, is all too familiar with this challenge.
Since 2016, the company has leveraged machine learning to monitor, identify, and proactively resolve IT infrastructure issues, and is now tackling the current problem of AI model reliability through an AI agent solution capable of everything from detection and diagnosis to remediation and prevention.
Founded by CEO Helen Gu, a computer science professor at North Carolina State University with previous experience at IBM and Google, the company recently secured $15 million in a Series B financing round led by Yu Galaxy, as revealed exclusively by TechCrunch.
According to Gu, the most pressing issue in the industry today is not merely monitoring and diagnosing failures in AI models; it’s understanding the operation of the entire tech stack now that AI has been integrated.
“To diagnose these AI model issues, you need to monitor and analyze the data, the model, and the infrastructure collectively,” Gu informed TechCrunch. “It’s not solely a model problem or a data problem; it’s a combination of factors. Sometimes, it can simply be your infrastructure.”
Gu illustrated this with a real-world example: one of InsightFinder’s clients, a prominent U.S. credit card company, detected drift in one of its fraud detection models. Thanks to InsightFinder’s extensive monitoring of the company’s infrastructure, it was able to pinpoint that the drift stemmed from outdated cache in various server nodes.
“The greatest misconception is that AI observability is confined to LLM evaluation during development and testing phases. In reality, a robust AI observability platform should offer end-to-end feedback loop support across development, evaluation, and production stages,” she asserted.
InsightFinder’s latest product, named Autonomous Reliability Insights, achieves this by employing a blend of unsupervised machine learning, proprietary large and small language models, predictive AI, and causal inference. According to Gu, this foundational layer is data agnostic, allowing the system to ingest and analyze entire data streams to extract signals that can be correlated and cross-validated to identify root causes.
The observability market is now crowded with competitors vying for a piece of the emerging sector created by the rise of AI tools. Nearly a decade into its journey, InsightFinder is up against significant players like Grafana Labs, Fiddler, Datadog, Dynatrace, New Relic, and BigPanda, all of whom are developing capabilities to tackle the new challenges posed by AI tools.
However, Gu remains unfazed. In fact, she believes that InsightFinder’s expertise, experience, and adaptability form a substantial competitive advantage. “We actually rarely lose [customers] to anyone so far… This is about the insights, right? The issue is that many data scientists understand AI, but they lack an understanding of the underlying systems. Conversely, many SRE [site reliability engineering] developers comprehend the system but not the AI… They often overlook the intrinsic relationships.”
Currently, InsightFinder’s customer base includes UBS, NBCUniversal, Lenovo, Dell, Google Cloud, and Comcast. Gu attributes their success to a decade spent understanding the needs of large enterprise clients.
“It has come down to collaborating with our Fortune 50 clients to refine and grasp the enterprise environment requirements necessary for deploying these types of models,” she mentioned. “We have been partnering with Dell to implement our AI systems globally for some of our largest clientele. This is not something you can simply apply foundational AI to machine data.”
Gu noted that the company’s revenue stream is “strong,” having increased “over threefold” in the previous year. Remarkably, she pointed out that InsightFinder was not actively seeking to raise this Series B, but investors approached following a seven-figure contract with a Fortune 50 company within three months.
The new capital will allow InsightFinder to make its initial sales and marketing hires, expand its team of fewer than 30 staff members, and invest in its go-to-market strategy. To date, the company has raised a total of $35 million.
