When we first started Storage Made Easy (or SMEStorage as it was in its original incarnation) we had a vision of software that provided a unification hub for corporate data, a single place where companies could manage content and set common policies for their company data..
Our vision embraced multicloud early, in fact way before multicloud was in-vogue the company’s tagline was ‘SMEStorage the multicloud company’.
The very first Cloud connector we added was Amazon S3, way before Amazon Web Services was the behemouth it is today.
In hindsight the companies name, SMEStorage and subsequently Storage Made Easy did not do the product justice. With prospects we joked that we don’t provide data storage but we do make the storage they have easier. In turn our prospects and customers affectionately referred to us as ‘SME’ and the product evolved into its own identity, The Enterprise File Fabric.
From the onset we mined content metadata as this was (and is) our secret sauce for working with the various storage solutions, be they file or object based. Over time the metadata collected evolved from simple file or object metadata (name, date, size, location) to also include content metadata (file or object content indexes). Today this includes integrations with AI based content metadata enhancement services such as Google Vision.
Whereas metadata was originally a means to an end for connecting and working with the (over 60) on-premises and on-cloud storage solutions that the File Fabric supports. today it provides the underpinning for services such as enterprise content search (a single search across all multicloud data sets) and content discovery (PHI, PII – content detection), data classifications and data catalogues.
Digital data is exploding and data is being stored further from the corporate perimeter than it ever has before, and across more siloed services than ever before. Companies don’t appear to be committing to a single cloud vendor and cheap on-premises object storage coupled with compliance regimes and fear of data breaches means organizations are adopting a hybrid approach to data storage.
This presents challenges for companies particularly with regards to how they corral this rapidly growing distributed data set, how they secure it, make sure its compliant (particularly with ever increasing legislation) and ensure it is available, usable and productive for their employees, and their business.
New technology innovations such as AI / deep learning are focusing companies on data strategy, even if they dont know exactly what that strategy should be right now – they know that they don’t want get left behind. We see this behaviour across a wide variety of industry verticals.
This brings us neatly back to metadata. As the File Fabric mines (multicloud) metadata, it is available to be used for the good of the organization in all sorts of ways which does not necessarily involve full blown AI or deep learning. This may involve using metadata to be smart about creating an Active Archive across two or more storage instances, aiding in the recovery of Ransomware attacks, or for Data Automation in which certain metadata event triggers cause an action to occur, such as tiering data (using the File Fabric’s M-Stream file acceleration feature). Internally we think of these as ‘smart correlations’ and we believe harnessing these can put a lot of power in the hands of companies with regards to how they put their disparate data to work for them.
As we move forward deep learning algorithms are introducing the possibility of being even smarter about finding patterns in the metadata between disparately stored data. This could be as diverse as finding files that are familial, or finding patterns that enable past behaviour to to predict future behaviour, all of which will be an enabler for a smarter document driven enterprise.