• by Tyler Loechner , Staff Writer @mp_tyler, August 11, 2015

Nearly two-and-a-half years ago, Real-Time Dailyspoke with data scientist Saed Sayad about the “predictive modeling” platform, called “Real Time Learning Machine,” he helped create for AdTheorent. 

Now with Rhostam, a new startup company devoted to bringing better tech to ad tech-ers, Sayad says he has built an even better platform, called Xarang.

Xarang is a predictive modeling technology Sayad asserts is the next generation of what he helped build at AdTheorent. It includes new features, such as the ability to A/B test, factor in market research analysis, optimize for price and more.

There’s a lot of fancy jargon -- what will all the “predictive modeling” and “optimizing” and “machine learning” and “real-time” lingo being tossed around -- but the general principle behind it all is simple to understand: Math, Sayad believes, can solve your programmatic advertising problems.

A marketer feeds the Xarang platform with first- and third-party data, explains Sayad, and the platform makes sense of it all. Perhaps the data indicates that you should shift 5% more of your budget toward mobile, or that your target audience would be more engaged if you reached them at a different time of day.

Sayad’s vision -- and hope, with his new Xarang technology -- is that these decision are made without any human intervention. That if the data indicates that a tweak should be made, the platform will automatically make the tweak.

“The goal is to provide ‘real-time learning machine’ technology to all marketers and agencies through a public or private demand-side platform (DSP),” comments Sayad. The company will license Xarang to buying platforms -- Rhostam itself does not provide a buying platform.

The technology will attempt to bring to life an “automate everything” future. It promotes a very hands-off approach, which actually goes against the grain of some of the ironic meta-trends happening in the programmatic industry.

In general, buyers and sellers are attempting to get more hands-on with programmatic technologies. For example, some brands are taking the tech in-house, while private exchanges and programmatic direct platforms are on the rise. Even Google’s decision to yank YouTube inventory from the DoubleClick AdExchange is an exercise in control. 

But, at least in theory, it’s possible for none of this micromanagement to be necessary. That’s the type of programmatic world Sayad is attempting to create -- one wherein disputes, such as trust issues between brands and agencies, are settled by “who has the better algorithm?” rather than “which partner is the most transparent?”

Simply letting the math be the great equalizer might be programmatic’s pie in the sky, but some are working toward it nonetheless.