Garbage in, garbage out. That’s long been the issue for successful data analytics, andit’s an even bigger issue today, giventhe rapid pace of investment inartificial intelligence (AI).
About80 percent of enterprises report they are investing in relatedAI technologies in some capacity and 30 percent plan onexpanding their investments, researchshows . Businesses expect AI to help keep them aheadof the competition.
Butrisk abounds. How can companies break the cycle of garbage in, garbage out andassure that insights derived from AIare sound?
Afterall, any information system―AI or other―is only as good as thequality of its data. Information systemsarelargely reliant on user input and entries need to be correct and credible.AI uses computer algorithms to replicate the humanability to learn and makepredictions, and AI software finds patterns and make inferences from largequantities of data. That’swhy the success of any AI effort will start with thedata. Good data in means good intelligence out. This requires thatenterprises:Ingest data quickly andseamlessly. Old and messed up data will result in misinformed intelligence.Enterprises needto ingest data quickly when it is fresh, and that means as itis. This is not an easy task. First, the volume of availabledata is exploding: IDC forecasts that by 2025, the global data-sphere will grow to 163ZB (i.e., atrillion gigabytes).That’s 10 times the 16.1ZB of data generated in 2016.Second, much of that data growth will be in “unstructured”data, such as videosand social media. That’s a problem for many enterprises because “unstructureddata doesn’teasily lend itself to older models of data storage and analysis,”IDCreports. Nor will such dataeasily lend itself to AIengines. Enterprises need to deploy technologies thatenable the rapid ingestion of data, including unstructured, sothat AI enginesanalyze the whole picture and make full use of data to better serve customersandanticipatetheirneeds. For data to be a winningcompetitive edge, it needs to be processed in real-time―not after never endingcycles of data normalization. Require data-levelsecurity. Security at the data level ensures the safety and integrity of the datawithin AIalgorithms. Data quality and security policies should be crafted atthe data level and be based off the metadata, whichis the data about the data.This guarantees that no matter where the data comes from or where ittravels to, securitypolicies will be with the data. It also means that, ifdata is changed, those changes are accounted for downstream andthe AI canadjust. The explosion of data sources―driven by such forces as the internet of things (IoT) and growth of mobiletechnologies―makes data-level security ever morecritical. Driveteamwork/collaboration. With a consistent view of data across organizations anddepartments, AI engines willbe more productive because only the best, mosttailored data will be accessed for analysis. Without teamwork andcollaboration, different departments may not know the full scope of availabledata. Teamwork and collaboration willbreak down data and department silos sothat only the right data at the right time is fed into the AI engine, and sothat data can be found more easily. It is also essential to have a consistentview of data that can be used across thelarger organization even if a slice ofthe data that a particular department needs is unique. AI Empowers
No doubt, AI has the capacity todisrupt all areas of a business and improve business performance. Marketresearcher Vanson Bourne , after conductingonline surveys of 260 senior IT andbusiness leaders last year, found thatcompaniesexpect a $2.87 return on investment over 10 years for AIinvestments.
To get there, enterprises will needthe full value of the data going into the AI engine. They need to ingest dataquickly,use that data, review what data is useful and then get more of thatuseful data. AI will enable enterprises to do thismore quickly, as long as thedata and the data management infrastructure is in good shape.