Salesforce draws on radar logic to push “operational intelligence” for real-time enterprise decisioning

In a long-ago quest to stop aircraft during the Battle of Britain, physicist Robert Watson-Watt’s team couldn’t build a “death ray,” but it produced radar instead. The lesson, Salesforce argues, is that the most valuable breakthrough is often not the raw detection, but the layer that turns indicates into a picture people can act on quickly. Today, the company says many enterprises live in a similar gap between having continuous streams of operational data and getting timely insight that prevents outages, reduces attrition, or protects key accounts.

Salesforce frames the problem as an overwhelming volume of information. Enterprises track error rates, sales pipeline movement, contact center telemetry, and even developer activity, with answers to many questions “somewhere” in millions of data points generated each hour. Yet the insight that could have changed what happened often appears after the decision window closes, leaving teams to react rather than prevent.

The core idea behind Salesforce’s push is what it calls “operational intelligence.” The concept is described as AI that reasons over a business’s live data continuously, rather than waiting for dashboards to summarize what already occurred. In the company’s telling, the system would behave more like radar operators scanning a sky in real time, delivering lead time so humans can apply judgment at the right moment.

Salesforce says two capabilities power this approach. First is Deep Insights, described as an AI reasoning system for structured business data. Deep Insights is said to trace problems back to root causes and recommend next steps, which the company positions as a bridge between data and decisions.

Second is Moirai, described as a time-series forecasting model that predicts demand and flags anomalies with business context already built in. In the company’s example framing, that context can include a product launch, a holiday season, or a wave of contract renewals, which helps forecasts align with how business events affect behavior.

Salesforce adds that parts of this approach are already being used inside its own operations. For Engineering 360, the company says the system tracks 90 billion records across dozens of internal systems. It then applies Deep Insights to identify patterns humans are not built to spot at that scale.

In remarks quoted in the release, Alex Hu, Salesforce’s EVP of Engineering Strategy, said Deep Insights gives Engineering 360 capabilities “no human would be able to accomplish at scale,” including “unprecedented pattern recognition” and “correlation across datasets” that would be impractical to do manually, alongside reasoning meant to explain what the data means.

The company also points to work in its Contact Center and Voice organization, applying forecasting and reasoning to reshape staffing and demand planning before problems hit the floor. Salesforce describes SVP Gautam Vasudev as discussing how these methods are intended to help the business anticipate workload shifts earlier, rather than treat staffing and demand as a reactive exercise.

Salesforce’s longer-term view is that operational intelligence should extend beyond engineering and support into sales pipeline health, marketing campaign performance, and supply chain shortage detection. The through line is the same: places where data volume has outpaced any human’s ability to watch continuously, and where the missing element is a layer that makes live indicates legible.

The release does not provide performance benchmarks, timelines for broader product availability, or detailed information about how Deep Insights and Moirai are integrated into any specific customer-facing offerings. It also does not disclose the exact scope of customer use cases, the models’ evaluation methodology, or how the company manages risks like incorrect recommendations. For now, The announcement is directionally clear, even if measurable outcomes remain unquantified in the public description.

Going forward, investors and customers will likely focus on whether Salesforce can translate the “radar” framing into repeatable results: faster root-cause discovery, earlier demand and anomaly detection, and demonstrable improvements in operational metrics. Watching for additional disclosures around deployments, model accuracy, and how “explanations” translate into faster decisions could indicate whether operational intelligence becomes a practical differentiator across enterprise functions.