AI-Powered Observability: From Reactive to Predictive

If there’s one thing clear from our AI-powered observability webinar, it’s that observability has officially graduated from a “nice-to-have” to a business-critical discipline, and AI is helping lead that charge.

Our webinar brought together guest speaker Stephen Elliott, Group VP at IDC, and Ranbir Chawla, former SVP of Engineering at RB Global, for an hour of insights that mixed data, experience, and hard-won lessons from the trenches. As moderator Chuck Czerkawski from Honeycomb put it, “We’re here to help customers achieve business value and realization through Honeycomb and observability.”

From reactive to predictive: the observability maturity curve

Stephen kicked off the discussion by sharing findings from IDC’s recent study, The Business Value of the Honeycomb Observability Platform. The common theme? Everyone wants to move “from reactive to proactive, and then to predictive and preventative.”

He explained that the best-performing teams are “thinking about their existing workflows, how they make decisions, and the data upon which they make decisions.” In other words, it’s not just about collecting telemetry, it’s about turning that data into foresight. And the results speak volumes. According to IDC’s interviews, Honeycomb customers saw:

• 79% less time to respond and remediate issues
• 73% faster mean time to recovery
• 45% fewer unplanned outages

As Stephen put it, “Honeycomb has prevented issues from happening because teams are catching things early now.”

The power of ‘system signal quality’

One of the most powerful concepts from the discussion was something Stephen credited to Ranbir: system signal quality. “This is about the quality of the information and data you’re collecting, and how that drives a more resilient architecture,” Stephen said.

Ranbir’s story brought it to life. During a massive digital transformation at RB Global, he said, “Things were so complex we were looking for the right observability tools. I had worked with Honeycomb very successfully before, so we brought them in as a partner.”

The payoff was instant clarity. “We had the tools to immediately get on a screen and share: ‘You told us it’s taking more than three seconds. I just found it,’” he recalled. Using tracing, even non-technical colleagues could spot issues: “A product manager looks and says, ‘Oh my God, what’s that big black line? That’s where our problem is.’ That’s the beauty of these visual tools. They make everyone care about why this works.”

Connecting code to customers

One of the most compelling parts of the fireside chat was when Stephen and Ranbir tackled the age-old question: How do you connect developer work to business value?

Stephen didn’t mince words: “We’re great at technology metrics, but the language of the business is finance. Every developer has a role in driving revenues and customer experience.”

Ranbir took it a step further, arguing that engineering leaders must act as translators. “The deeper you understand the business and engender trust, the better your career, and the happier your teams. You can bridge those conversations and stop random questions and stress.”

It’s that bridge—between code deploys and customer satisfaction—that observability makes tangible. As Ranbir said, “When something’s wrong, we can show whether it’s our code, a partner’s system, or even a CDN issue. That’s power.”

Making developers business-fluent

Both speakers agreed that business fluency starts with intentionality. “It always starts with a conversation about what we’re building and why we’re building it,” said Ranbir.

He urged teams to add contextual telemetry: “Think about tagging signals properly. You can have a business SLO—‘We need to get this much money in’—and a system SLO underneath it.”

Stephen echoed the sentiment: “A good service level indicator can be expressed meaningfully in a sentence that all stakeholders can understand.” That clarity helps teams eliminate finger-pointing and align on what success looks like.

The AI layer: promise and complexity

Finally, the group turned to the role of AI. Stephen predicted that “every organization will have a massive number of agents—AI assistants for root cause analysis, prediction, and co-generation.”

But Ranbir, self-proclaimed “grumpy old man on a park bench,” offered a grounded take: “There’s a lot of math available to us. Combine your Honeycomb telemetry with your chaos testing data and you can start predicting issues before they happen.” He also sees opportunity in using AI for toil reduction by automating repetitive engineering work and freeing humans to focus on innovation. “Get LLMs focused on toil, not replacing devs. That’s where the magic is.”

Final thoughts

Observability isn’t just about seeing what’s broken. It’s about understanding how technology decisions ripple into customer experience and revenue.

As Stephen summarized, “It’s not rocket science. Understand what your business values, and connect it to what you already measure.”

And, as Ranbir put it, “If you can prove your point with data, you’ll live a long time as a team.”