Register for "Diverse, Integrated, and Real Time: Delivering the Right Data for AI/ML Success" - Tuesday, June 18,1:00 pm EST

Decision AI and the Opportunity for Smarter, Faster Actions

ABSTRACT: Decision AI is an emerging discipline that enables faster, smarter operational decisions by combining decision intelligence and contextual intelligence.

This blog is sponsored by Diwo.

Enterprises struggle to make operational decisions because—paradoxically—they drown in analytics.

Analysts and business managers use traditional business intelligence (BI) tools to understand what happened, data science tools to predict what will happen, and both types of tools to understand why. But they have too many charts to synthesize, too many data points to correlate, and too little context about the results of their last decision. Overwhelmed, they default to intuition and gut instinct. This circuitous process results in bad decisions, late decisions, or no decisions. Business performance suffers.

Enter Decision AI

However, enterprises have a new analytical option to make more timely, accurate decisions that drive the right actions. Decision AI is an emerging discipline that uses artificial intelligence (AI), automation, and metadata to improve the decision cycle for operations. Like other technologies, decision AI prepares data for analytics and helps users visualize analytical outputs. But the real innovation lies in its decision intelligence and contextual intelligence.

Decision intelligence detects what happened and analyzes why it happened. Then it predicts what will happen, prescribes what to do, and automates the resulting decision and action. Decision intelligence uses visualization, augmented analytics, and machine learning (ML) to help make decisions, then workflow orchestration to automate the resulting action. Vendors such as Diwo, Noodle, and Aera provide tools that support decision intelligence, including recommendations.

Contextual intelligence helps make those decisions actionable and specific, and more so over time because it learns. It assesses outcomes, compares them to expectations, then refines algorithms to improve the outcomes of future decisions. Contextual intelligence uses machine learning to perform these tasks and visualization to present evidence along the way. It overlaps with decision intelligence because it also helps prescribe what to do and decide how to act. Diwo's Decision Intelligence Platform combines contextual intelligence and decision intelligence in this way.

Decision AI seeks to drive smarter, faster, and more adaptable business thinking than traditional BI reports and dashboards. Designed and implemented well, it has the potential to surpass human limitations in the following ways.

  • Smarter. Decision AI examines many data points and combinations, then isolates the most relevant ones. This helps replace human shortcuts and hunches with a comprehensive, evidence-based conclusion about what factors really matter.
  • Faster. Decision AI automates the cycle of analysis, decision, and action. This helps replace human hesitation and second guesses with rapid processing of the facts.
  • More adaptable. Decision AI automatically catalogs decisions and their outcomes, then performs a post mortem to improve future decisions. This helps replace excuses and bad habits with logical ideas to avoid repeat mistakes.

Case Study

So how might this vision play out in reality? Suppose a toy manufacturer needs to streamline its supply chain to meet forecasts for holiday shopping. The supply chain manager reviews shipment orders and delivery schedules across 12 suppliers in four countries. The decision AI platform guides them through a series of logical steps.

Decision intelligence. Alerted by the platform, this manager detects that one toy supplier just pushed back its shipment dates. Prompted by popup windows, they analyze why this happened and identify the root cause: delays from one of that supplier’s parts factories, which has a history of missing deadlines. The decision AI platform predicts that the factory supplier will delay its shipments further based on historical trends.

Next, the platform prescribes a decision: cancel the order with that supplier, and instead order those parts from another supplier with a better history of shipping on time. The supply chain manager validates this thinking with the data analyst, then decides to take the recommended action. This kicks off an automated workflow that makes the change.

Contextual intelligence. Now things get interesting. The decision AI platform assesses the outcome of that decision over time and compares it to expectations, noting that the new supplier does meet its shipping deadlines. It logs this success within a catalog of all relevant decisions and outcomes for review by the supply chain manager. In addition, the decision AI platform suggests ways to refine future decisions like these. In this case, it prompts the manager to add real-time alerting of changes to factory schedules. The real-time alerts give them advance warning of shipment delays, enabling faster and more proactive responses.

That’s the vision of decision AI. To realize the vision in practice, enterprises must strike the right balance between machine logic and human judgment. They must focus on a narrow operational domain that involves time-consuming, tactical, and repetitive decisions. They must navigate risks that include hyper competitive markets, geopolitical shocks, and economic uncertainty.

All easier said than done! We will explore these opportunities and challenges further in my upcoming report and webinar.

Kevin Petrie

Kevin is the VP of Research at BARC US, where he writes and speaks about the intersection of AI, analytics, and data management. For nearly three decades Kevin has deciphered...

More About Kevin Petrie