Decisions, Decisions… Next Steps For Practical Applied AI

Numerous companies specialize in robotic process automation (RPA), offering bots to automate repetitive tasks—UiPath, Automation Anywhere, Blue Prism, and others. Meanwhile, process mining, orchestration, and intelligent document processing specialists like Celonis, IBM, Abbyy, and Aris occupy a different niche. Then there are AI-driven supply chain planning and forecasting experts, including ERP giants SAP and Oracle, alongside Blue Yonder and Coupa. A separate category exists for digital twin specialists utilizing IoT AI toolsets, such as o9 Solutions, Bentley Systems, Siemens, and General Electric. Finally, we have general-purpose AI providers like IBM Watsonx, OpenAI, Anthropic, Microsoft, and Google.

The next frontier for many of these vendors is the direct application of AI within business processes. Demonstrating AI's practical, quantifiable impact on business decisions is crucial, as AI alone remains a passive resource.

Prioritizing Simplicity

While discussions around autonomous AI decision-making are prevalent, the field remains somewhat nebulous. Clear-cut decision management tools for supply chain, finance, and operations could prove highly beneficial.

This context helps explain Aera Technology's focus on "decision intelligence"—software designed to recommend actions within business processes. Aera Decision Cloud predicts supply shortages, suggests shipment rerouting, and automates inventory replenishment without human intervention.

"Decision intelligence is AI for decisions; it optimizes decision-making across a business," says Fred Laluyaux, CEO and co-founder of Aera Technology. "As decisions become more frequent and complex, organizations need faster, data-driven decisions. By recommending and automating actions, decision intelligence accelerates the decision cycle, improving efficiency, reducing costs, enhancing customer service, and facilitating responsiveness to change."

Laluyaux suggests decision intelligence is fundamentally altering how large enterprises work and make decisions, potentially reshaping business operations and creating new roles focused on managing this technology. Scaling these tools requires addressing bias and capturing the business context of each recommendation and decision. Real-time application to complex scenarios represents a new level of automation.

A Retrospective on Decision-Making

"Companies make countless decisions daily impacting costs, efficiency, and customer satisfaction. Often, there's no record explaining the rationale behind these decisions," Laluyaux explains. "This lack of insight perpetuates inefficiencies. Meanwhile, the pace and complexity of decisions continue to rise."

Aera's decision intelligence technology documents each decision and its reasoning. It goes beyond simple problem detection, aggregating data from various sources—structured databases, unstructured text, IoT devices, cloud applications—to provide a holistic view. The technology enables the creation of "decision flows" (schematics illustrating decision components) and integrates AI models while allowing custom interface development.

Gartner predicts that by 2026, 75% of Global 500 companies will utilize decision intelligence practices, including decision logging. Other forecasts suggest that by 2028, 15% of daily work decisions will be automated via agentic AI.

The Price of Inaction

"Failing to capture decision rationales leads to lost learning. Traditional back-testing reveals trends but not the 'why' behind human choices," Laluyaux states. "For example, a planner might override an AI forecast due to a known late customer order—information unavailable to the AI. Without recording the override reason, the model cannot learn, failing to anticipate similar situations."

Similarly, planners might reject AI suggestions due to perceived risk, distorting data through manual adjustments.

Aera serves clients including Unilever, Merck, ExxonMobil, Mars, Kraft Heinz, and Dell. But does decision intelligence unify the AI market and illuminate the path forward? The answer is multifaceted.

Aera's pre-built industry models for functions like supply chain management offer an advantage over basic RPA bots. Its closed-loop automation creates autonomous decision-making cycles based on approved protocols, exceeding the capabilities of basic analytics or task execution tools.

Limitations of Specialization?

However, specialization entails trade-offs. Aera's focus on specific enterprise functions limits its intelligence engine compared to IBM's Watsonx, which can process a broader range of information.

Aera also relies on businesses embracing autonomous decision-making, a concept not yet universally adopted.

Furthermore, Aera operates in a crowded market dominated by more established platforms with wider implementation and comprehensive partner ecosystems. In a complex technology landscape, this is a significant factor, regardless of AI capabilities.

The above is the detailed content of Decisions, Decisions… Next Steps For Practical Applied AI. For more information, please follow other related articles on the PHP Chinese website!