Decision Intelligence for Operations: The Next System Layer in Enterprise Software

This week I'm at HumanX talking with founders, investors, and operators about something I've been thinking about for a long time.

Over the past twenty years, enterprises have invested billions of dollars digitizing their operations. They built systems to capture data from machines, processes, and enterprise workflows. Today most industrial organizations have more operational data than ever before.

But something important is still missing. The systems they built are very good at telling them what happened, but they are much less capable of helping teams decide what to do next.

That gap is what led us to start SteelTree. The core idea behind the company is simple:

• Enterprises spent the last twenty years digitizing operational data
• The next decade will be about digitizing operational decisions themselves

The Reality of Operational Decisions

Across industrial environments, frontline teams make thousands of decisions every day, like:

• How to respond to abnormal conditions
• Whether to adjust production
• Which maintenance issue to prioritize
• How to handle a quality deviation
• When to intervene for safety

These decisions determine reliability, safety, and efficiency in real time, yet most of them share the same characteristics:

• They rely on fragmented data
• They depend heavily on individual experience
• They are rarely captured
• They almost never improve systematically over time

In many organizations, the most valuable operational knowledge still lives in people rather than in systems. When experienced operators leave or teams rotate, much of that knowledge leaves with them.

The Missing System

That gap is what led us to build SteelTree, a Decision Intelligence system for operations, designed specifically for industrial environments. Instead of focusing only on capturing data or generating reports, the system focuses on the moment where operations are won or lost, the decision itself, and builds operational context by connecting:

• Operational signals
• Human judgment and experience
• Historical outcomes
• Real-time operational events
• The actions teams ultimately take

The goal is not simply to analyze data. It is to help teams understand what is happening, decide what to do next, and continuously improve how operations run. Over time, every decision captured by the system contributes to a growing operational knowledge base unique to that organization.

Why Now?

For a long time, building a system like this simply wasn't practical. Operational environments generate messy, fragmented signals across dozens of systems. Interpreting those signals in context required human expertise.

Recent advances in AI now make it possible to interpret operational signals and combine them with human input in real time. This reality finally allows a new layer to emerge in the enterprise stack, a system designed specifically to support operational decision-making.

How SteelTree Fits into the Enterprise Stack

Most operational teams already have enormous amounts of data across dashboards, reports, and monitoring systems. But interpreting that data quickly in context often requires significant effort.

The first value SteelTree provides is simply helping teams understand what is happening. Instead of searching through multiple systems or building reports, users can ask straightforward questions:

• Why did output drop this shift?
• Where are we at risk right now?
• What changed in the last hour?

The system interprets signals across systems and presents a clear explanation, and once teams trust that the system understands their data, its role naturally expands. SteelTree begins to highlight emerging issues, patterns, or anomalies that may require attention. From there it can suggest actions based on how similar situations were handled previously and eventually anticipate conditions before they become problems, always working alongside the team.

In practice, teams using SteelTree move naturally through a progression:

Understanding -> Awareness -> Guidance -> Prediction

Most traditional systems helped organizations reach the first step. SteelTree is designed to make that step dramatically easier and then carry the process forward into the decision layer of operations.

A Personal Journey That Led Here

My own path to SteelTree started long before the current wave of AI. Over the years I've worked closely with organizations running large-scale operations and building software used by frontline teams, and one pattern kept appearing.

Enterprises made enormous progress in digitizing operational data. Dashboards helped teams understand what had happened, analytics highlighted patterns and emerging issues, and reports summarized performance across the organization.

These capabilities were incredibly valuable because they gave teams a level of visibility into operations that simply didn't exist before. But reaching that understanding came at a cost.

Building dashboards, analytics pipelines, and reporting systems typically required complex data infrastructure, specialized skills, long implementation cycles, and significant investment. Even then, the information was often difficult for operational teams to interpret quickly in the moment.

In many environments, enormous effort was spent simply getting to a basic understanding of what the data was saying, and once that understanding was reached, another gap appeared.

When an issue surfaced, when conditions changed, when a trade-off had to be made, the most important question remained the same: what should we do next?

At that moment, the system usually stopped, and the decision relied almost entirely on human interpretation and the hope that the right skills were available to make the correct decision.

SteelTree is designed to address both challenges. It dramatically reduces the complexity required to understand operational data. Instead of building dashboards and reports, teams can interact directly with their data and see patterns, explanations, and visualizations emerge immediately.

But SteelTree does not stop at understanding. As the system observes patterns and decisions over time, it begins to highlight emerging issues, suggest potential actions based on similar situations, and eventually anticipate conditions before they occur.

What Excites Me About This Moment

The idea of Decision Intelligence is still emerging, but it is becoming increasingly clear that a new system layer is forming between operational data and operational execution.

If we get this right, the impact is significant. Organizations will not only see what is happening in their operations, they will continuously learn how to run them better.

• Decisions will become more consistent
• Operational knowledge will compound rather than disappear
• Teams will be able to move faster and with more confidence

Conversations at HumanX

If you're attending HumanX this week, I'd love to talk.

We're looking to work with companies, large and small, that want to explore new ways to interpret operational data and turn it into faster, more confident decisions.

We're also always interested in conversations with investors who believe the next generation of enterprise systems will be built around context-aware decisions, not just data.