Early warning system to warn business managers about adverse business outcomes due to unplanned changes in data sets (data drifts).
Data observability has emerged as a category in the data integration and intelligence software market to continuously monitor and test data for drift, shift, and anomalies, and notify those who need to know.”, said Stewart Bond, director, Data Integration and Intelligence software research at IDC
Read More: Over 4000 Brands Participated In Onion Global’s 2021 Double 11 Shopping Festival…
iMaaS’ comprehensive set of features warn business managers and data scientists about adverse business outcomes that could be caused by data drifts before they actually happen. Data drifts occur because of unplanned or unaccounted for events in the business ecosystem, such as onset of adverse weather, or the arrival of a pandemic. The platform allows business managers to plan for such adverse changes, across diverse data sets. iMaaS delivers warnings about the impact of drift and identification of the root causes, thus allowing decision makers to take pre-emptive action. For example, according to the Chief Data Officer of a leading insurance company, an alert about a potential increase in accident rates in certain zip codes due to unseasonal cold weather, would be greatly helpful to re-price policies appropriately, as well as set aside the right capital amounts to pay off claims, without which the insurance company could suffer losses from claims overages.
“Data drift and Prediction score drifts are insidious problems as businesses increasingly deploy AI-powered solutions that depend on the quality of data that is used to build the AI models for mission-critical applications” said Dr. Valeria Sadovkyh, a technology consultant for a prominent consulting company. “A very prominent online real-estate firm for prospective buyers recently shut down its house-flipping division after the algorithm proved unable to forecast housing prices with sufficient accuracy. While the models were built with historical data sets, the Covid pandemic introduced data drift which rendered the AI solutions invalid. Learning examples failed because the distribution of data from the real world shifted from the original data sets. Constant monitoring of data and predictions is thus an imperative for the success of AI/ML solutions deployed in production.”
Read More: SalesTechStar Interview With Alina Trigubenko, CEO And Founder, Profi
“Drift Detection is a largely unsolved problem in the industry until now” says Shiraz Zaman, Head of AI Platform Engineering at a leading ride-hailing company. ScoreData has uniquely solved this problem, using patented technologies to monitor, measure, control and act, to allow businesses to use AI not only to improve data quality but to also continuously monitor and test data for a wide variety of changes through pre-production and post-production launches and operations of business critical software applications.
“87% of CXOs are aiming to become a data-intelligent enterprise, but data is increasingly more distributed, diverse and dynamic, making it difficult to govern and control. Data is also continuous – continuously moving and changing, requiring data intelligence and governance functions to also be continuous and synchronous with data. Data observability has emerged as a category in the data integration and intelligence software market to continuously monitor and test data for drift, shift, and anomalies, and notify those who need to know.”, said Stewart Bond, director, Data Integration and Intelligence software research at IDC.
