Artificial Intelligence (AI) is starting to make big waves in business IT. From chatbots to virtual assistants, ‘smart’ automation that learns on the job to predictive analytics which can forecast future trends with unerring accuracy, AI is shifting the boundaries of what digital technology is capable of.
With AI going mainstream, the new normal is – why get a person to do it when a machine will do?
According to Gartner, AI adoption across commerce and industry has tripled since 2018. That kind of demand means the technology is front and centre of developers’ minds when it comes to designing future-facing solutions for clients. But particularly in the field of database development, AI poses some significant challenges.
The fact is that machines do not think like humans do. Instead of the creative and associative synthesis of concepts which lies at the heart of human reasoning, AI depends on the ability to process data – and massive volumes of it – to achieve outcomes that resemble (and in many cases outstrip) our own intelligence.
For AI disciplines like machine learning and predictive analytics, data handling and preparation is therefore a core priority. It can also be cumbersome and time-consuming – data specialists spend as much as 80% of their time focusing on these areas to ensure AI applications deliver the outcomes desired. This makes data preparation a key barrier to fast, efficient development and roll out of AI solutions.
Freeing DevOps Capacity
Enter IBM and its brand new InfoSphere Advanced Data Preparation solution. The idea is simple – to make the engineering of raw datasets ready for use in AI applications a less time-intensive process, freeing up DataOps capacity to focus on more added-value areas and removing a major bottleneck in the AI roll-out process.
InfoSphere is designed to sit on top of existing datasets and provide visualisation, tracking and cleaning tools. It can work with extremely large volumes of raw data – so-called data lakes – which an organisation may need to harness to make AI and advanced analytics applications viable. Transforming unstructured data into structured data is notoriously time consuming, so InfoSphere uses automation tools to slash the workloads involved.
One important feature of InfoSphere is that it makes use of an intuitive user dashboard which is intended to make the tools accessible to the non-data specialist. This means that instead of relying on data scientists to carry out all aspects of data preparation, ordinary business users can get involved in the process and take on some of the burden. As well as having specific applications in database development for AI, this also helps to upskill the general workforce in a hitherto highly technical and specialised area of IT which is increasingly being viewed as a core area of value generation for businesses of all types and sizes.
Ultimately, for businesses to enjoy the advantages on offer to operational efficiency, customer experience design and strategic decision making that AI and advanced data analytics offer, they have to be in a position to turn raw data feeds into manageable sets that can be used by the applications at their disposal. That is what InfoSphere promises to make a whole lot easier.