Enterprise tech startup Aera today unveiled a product aimed at helping businesses use machine intelligence to make decisions.

Company decision makers can ask Aera questions about their business using a voice interface or text search and get back proactive suggestions about what they should do to improve the outcome. For example, the system could answer a question about revenue by suggesting an inventory shift from one territory to another.

Those suggestions are based on a skill builder Aera provides that allows enterprise users to define workflows that run in response to certain questions. For example, companies can set up a revenue workflow that checks the system’s supply chain model and provides a course of action.

Aera is designed to connect predictive modeling capabilities with the ability to take action, which should make it easier for companies to go from insight to business outcomes. The system also changes its interface to fit each user, based on machine learning.

By using Aera, companies should be able to move from automating their transaction processing toward automating their decision-making, according to Fred Laluyaux, the startup’s CEO.

“Automating the way you make decisions, and the way you execute those decisions, this is critical,” he said. “Not at the transaction level, but the level above that.”

The key innovation behind Aera is its use of crawler programs that can trawl systems of record and send normalized business data back to the cloud in real time for further processing, while also writing new data into those systems. That means it’s possible for companies to collate data from a mainframe and an Oracle ERP system for generating insights, then take action by sending new data from the cloud into those legacy systems.

There’s no shortage of companies out to connect disparate enterprise systems and then attach AI to help businesses derive insights, but automatically collating data from a variety of systems and labeling it properly can prove challenging for many of them.

Aera’s core technology is based on expanding and rebranding the work of FusionOps, a company that created software for supply chain management. While that software was focused somewhat narrowly, Aera is built to serve a wider range of enterprises and includes features FusionOps didn’t have, like the ability to create custom workflows.

The FusionOps pedigree helps with customer acquisition, since Aera’s system is already being used by major enterprises like Merck and Columbia Sportswear.