o9's Chakri Gottemukkala on Becoming the Enterprise AI Model

The Editorial Team, o9

Published on: August 25, 2026

7 read min

Large enterprises have spent years trying to connect planning across functions in an effort to patch an all-too-familiar issue: Commercial teams work with one set of assumptions. Supply chain teams work with another. Finance, procurement, operations, and product functions each have their own processes, data, and decision cycles.

For Chakri Gottemukkala, Co-Founder and CEO of o9 Solutions, those silos remain one of the biggest sources of lost value in large organizations.

“The biggest value leakage is in the silos,” he said at aim10x Europe.

The problem has become more urgent as volatility increases and AI develops at speed. Businesses now need operating models that can respond faster, learn continuously, and bring intelligence closer to the people making decisions.

That is the thinking behind APEX, o9’s model for Agile, Adaptive, Autonomous Planning and Execution.

Learning from the simplicity of a small business

Gottemukkala traced the idea back to a simple story.

When Herman Lay was building what would eventually become Frito-Lay, he effectively operated a business without functional silos.

He could see what consumers were buying, speak directly with retailers, understand supply conditions, decide what to produce, adjust pricing, and learn from the results.

All those decisions were connected.

“He had perfect value chain visibility end to end,” Gottemukkala said. “He knew what was going on at the consumer end, he knew what was going on at the potatoes end, he knew everything that was going on in the internal operations.”

As companies grow, that simplicity disappears.

More markets, products, functions, systems, and layers of management make decision-making increasingly fragmented. Information travels through processes and organizational structures before reaching the people who need it.

The challenge is to recreate the connected decision-making of a small business at enterprise scale.

That idea became the foundation for o9’s Enterprise Digital Brain.

Connecting decisions across the enterprise

The Digital Brain was designed to connect planning across commercial, product, supply chain, and execution processes.

Long-range decisions can be linked with tactical plans and daily execution. A change in one part of the business can be evaluated against its impact elsewhere.

The aim is to reduce the value lost when functions optimize independently.

But Gottemukkala believes connecting plans is only part of the problem.

Large transformations themselves remain slow.

Organizations may spend years identifying problems, agreeing on an approach, securing funding, implementing technology, and driving adoption.

That pace is increasingly difficult to justify when value is already leaking from the business.

“If you’re losing $200 million per year EBITDA for a $10 billion company, why is it taking so long to even get a transformation program up and running?” he asked.

The question behind APEX is whether improvement can become continuous rather than dependent on periodic transformation programs.

Moving from agile to adaptive

APEX stands for Agile, Adaptive, Autonomous Planning and Execution.

The agile element builds on connected planning. It is about reducing silos and increasing the speed of decision-making.

Adaptive goes further.

Gottemukkala describes it as the ability of an organization to learn from performance, identify where value is leaking, and improve continuously.

“How are you constantly learning every day and improving versus waiting for a transformation program that takes forever and is too risky?” he said.

This requires organizations to understand the reasons behind performance gaps.

Most companies can see what happened. Inventory increased. Margins fell. Service declined. Demand missed the forecast.

The harder question is why.

That answer may sit across several functions. Poor master data, incentives, process design, capacity decisions, pricing, or supply constraints could all contribute.

Without a shared understanding of the cause, cross-functional alignment becomes much harder.

Diagnosing where value is leaking

Gottemukkala compares the future enterprise operating model to healthcare.

Doctors do not simply observe that a patient is unwell. They rely on a body of knowledge, diagnostics, and established approaches to understand the cause before determining how to respond.

He argues that enterprises need something similar.

“If you ask the question, why is excess inventory building up, why is margin leaking, why is growth not what it needs to be, it’s a very tough question to answer,” he said.

APEX is intended to help organizations identify these value leakages, understand their causes, decide what needs to change, and sustain the improvement.

That last point matters.

Gottemukkala noted that some organizations complete major transformation programs only to find themselves facing similar problems several years later.

The objective is therefore to build learning into the operating model itself.

Bringing AI and enterprise knowledge together

AI is central to this vision, but Gottemukkala argues that large language models alone are not enough.

An enterprise is an interconnected system of products, customers, suppliers, capacities, constraints, policies, and decisions. AI needs business context to understand those relationships.

o9’s approach combines what Gottemukkala describes as neural AI, including large language models, with symbolic AI captured through the Enterprise Knowledge Graph.

“It’s capturing the knowledge of how enterprises’ decision-making is connected across the value chain,” he said.

The combination is intended to make enterprise knowledge easier to access and apply.

Instead of relying only on planners to interact with sophisticated models, AI could help bring those models closer to frontline decision-makers.

Moving decision-making closer to the edge

Today, many planning systems are primarily used by planning teams.

Yet important business questions often originate elsewhere.

An account manager may see a new revenue opportunity. A procurement team may identify a raw-material risk. An executive may need to understand how a pricing decision affects margin and supply.

Those questions then travel through planning processes before the analysis comes back.

That creates delay.

Gottemukkala believes AI can reduce those layers by allowing frontline users to interact more directly with enterprise planning models.

“How do we move analysis and decision-making to the edges, reduce the layers, remove the latencies, remove the frictions?” he asked.

Consider a raw-material price increase.

Answering the business question may require understanding which finished products are affected, how margins will change, whether prices should rise, how customers might respond, and what alternative actions are available.

Today, that analysis may involve supply chain, procurement, commercial, finance, and executive teams.

In a more autonomous model, much of the underlying analysis could happen faster, while people remain focused on the decisions that require judgment.

Using post-game analysis to understand performance

One of the capabilities Gottemukkala highlighted is what o9 calls the performance post-game analyzer.

The idea comes from sport.

Teams review games after they happen to understand why actual performance differed from the plan. They study individual decisions and events so they can improve the next time.

Enterprises face similar gaps between planning and execution every day.

Demand differs from forecast. Inventory builds in unexpected places. Margins move. Suppliers miss commitments.

“Most people get the what,” Gottemukkala said. “Can we understand the why behind that plan versus execution deviation?”

If companies can identify those causes quickly, they can also decide what to change faster.

That learning then feeds into the next planning cycle.

Simulating the future before living it

APEX also includes the concept of business simulators.

Gottemukkala compares them with flight simulators, where pilots can rehearse difficult situations before encountering them in the real world.

Companies could apply a similar approach to business volatility.

Teams might simulate price increases, cost changes, supply constraints, or unexpected demand opportunities using models grounded in their own enterprise and industry data.

This could help organizations prepare teams for unfamiliar scenarios and align more quickly around future ways of working.

Rather than explaining a transformation through slides and process diagrams, employees could experience how decisions would work in the future model.

Building the capability to keep improving

Gottemukkala also sees a stronger role for internal centers of excellence.

If organizations depend on external vendors for every improvement, the pace of change remains constrained.

Internal capability gives companies more control over how quickly they develop new models, adjust processes, and respond to emerging opportunities.

“How do we get your organization to have a center of excellence that can drive the innovation in your company?” he asked. “You have the power to continuously improve.”

This reflects the broader idea behind APEX.

Transformation should become less episodic.

The long-term ambition is an operating model that can identify problems, learn from performance, apply new technology, and improve continuously.

Gottemukkala described the end state as a future where implementation itself becomes much lighter.

“The model is live on day one,” he said. “As you add more and more data to the model, your private data, your enterprise data, the model generates more value.”

As volatility increases and AI becomes more capable, enterprises will need to shorten the distance between signal, analysis, decision, and action.

The next step is to build operating models that can do that continuously, while learning from every decision they make.