When Sid Sachdeva was heading up supply and procurement at Reliant Energy, many efficiency companies pitched him, yet most lacked the type of industry applications that his team would have been interested in.

“There was no clear focus on something that was economically viable,” he said. He noted that most companies had built applications that targeted ratepayer-funded efficiency programs -- trying to help utilities better connect with their customers and helping utilities make better use of energy-efficiency dollars, which have to be spent no matter what.

Sachdeva left Reliant and founded Innowatts, which helps energy retailers, regulated utilities and end customers attack costs rather than just making better use of ratepayer efficiency dollars. “We optimize the cost of energy supply and forecast it at the individual customer data level,” he explained.

Innowatts built a machine-learning platform that leverages smart-grid data and weather or property data to build an individualized picture of end loads that disaggregates weather sensitivity.

Innowatts currently counts four retail energy providers and three utilities as customers since launching in summer 2014. The company started in regions with smart meters already deployed, but Sachdeva says the platform can now be used in utility territories without AMI. Innowatts claims it can cut the cost of serving customers by more than 20 percent.

Retailers such as Just Energy are using the platform to optimize supply cost at a more granular level, as well as to retain customers. Energy retailers can quantify the churn probability for each customer, said Sachdeva, and then use that information to alert customers to high bills once they know the probable shock threshold for each customer. By being more proactive about bill alerts, retailers can reduce churn.

For regulated utilities, there are also operational savings, such as reducing low-income subsidies by targeting the right low-income customers for energy-efficiency upgrades. The platform can also be used to target which customer is right for demand response, dynamic pricing or solar, applications that are also offered by competitors such as Opower and Bidgely.

Innowatts will find many competitors from efficiency and demand-side management firms on both the residential and commercial sides. Sachdeva acknowledges that some competitors offer similar applications, but the primary goal of helping retailers tackle supply costs across the total customer base is an application where he does not see much competition.

For one retailer, Innowatts was able to improve supply predictions by about 25 percent during a frigid January. Generally, the company is delivering about a 5 percent to 10 percent increase in energy margins in terms of dollars per megawatt-hour to its retail customers.

In many cases, the analytics from Innowatts are complementing other customer-facing software analytics. Earlier this year, for instance, Opower rebooked Sacramento Municipal Utility District for enterprise-wide digital engagement and energy efficiency, and SMUD has also chosen Innowatts’ platform to help manage customer costs.

The Houston-based company is focused on the North American market, especially the deregulated states, but has also been piloting in Europe and Asia. Innowatts has raised about $1 million to date from an undisclosed list of investors.