How ‘humanised machine learning’ platforms can help businesses make sense of their data
Mind Foundry, a technology spinout from the University of Oxford’s Machine Learning Research Group, has launched a cloud-based ‘humanised machine learning’ platform that allows anyone of any technical ability and in any size of organisation to unlock the full value of their data and make decisions on complex business issues without the need for data scientists.
Offering data analytics capabilities previously only available on a consultancy basis, the platform automates the building of machine learning models that can identify and visualise insights contained within data and recommend or automatically initiate an outcome.
Once the user has uploaded relevant data to the platform, Mind Foundry guides them through simple steps to build and deploy the optimum machine learning model for a business problem, which can then be integrated into a company’s website or business process. All within minutes or hours, rather than weeks or months.
Mind Foundry says its platform is so easy to use that by 2020 it aims have turned 1,000 business users into ‘citizen data scientists’ capable of analysing the huge amounts of data within their organisations to optimise business processes, improve day-to-day operations and enhance decision-making.
Here, Mind Foundry Director of Research Nathan Korda outlines some of the benefits of this new approach.
Businesses are already struggling to make use of the huge amounts of data available to them and, with worldwide data volumes predicted to grow by 61% between 2018 and 2025, according to IDC, they are only going to find it harder if they continue to rely on spreadsheets and simple data models that can only extract limited value from data.
Machine learning (ML) can give organisations greater insight into their data, enabling them to harness data to optimise business processes, improve day-to-day operations and inform decision-making. However, ML has traditionally required extensive resources, time and technical expertise.
Often, this involves the hiring of data scientists, which presents problems of its own: demand for this highly specialised skill currently outstrips supply and, too often, data scientists are remote from a business problem and not fully able to contextualise it and understand its full impact on operations.
What’s needed are ‘citizen data scientists’ – employees not operating in dedicated data science or analytics roles, who can use machine learning to explore their data and deploy models to unlock the value it holds, freeing up data scientists to focus on digital transformation projects.
Herein lies the value of Mind Foundry’s user-centric ML platform. By giving data owners in any size of organisation access to advanced machine learning technology without the need for significant investment or specialist training, it empowers them to master their own data and complete operations at scale.
A machine learning platform provides citizen data scientists with the capabilities required to prepare and visualise data and subsequently build, deploy and manage a suitable model, guiding users through the whole process, from suggesting actions to clean and format data to recommending the most suitable model for a data set.
This has real benefits in reducing the volume of mundane data preparation tasks involved in business processes that are repetitive and analyse data in a similar way on a routine basis, such as budget forecasting.
Instead of tying up senior management resources for several weeks to finalise budgets based on expected business outcomes, managers can use an intuitive machine learning platform to quickly identify and set up a model capable of being reused to revise budgets annually, dramatically cutting the time invested in this process going forward.
Crucially, these will still be operational and reusable by colleagues even after the citizen data scientist that created them has left the company.
Another benefit of a humanised machine learning platform is that ownership and control of a process remains with the user. Consider the case of a precision components manufacturer. Instead of relying on a data science team to identify patterns and areas for optimisation hidden within data produced by equipment sensors, an ML platform would enable the company’s machinery experts to input, cleanse and visualise this data in minutes and then select an appropriate data model to uncover previously unseen insights.
Business problem owners will always have a unique and intimate knowledge of a specific problem and its relevance to existing business priorities. With a humanised machine learning platform, they can directly identify and enhance the value of their data by quickly harnessing machine intelligence at scale.
Machine learning is excellent for risk assessment and management and for making data-driven judgement calls, but lacks the intuition and creativity required to contextualise and problemsolve. Use it to handle labour-intensive and repetitive tasks, such as data cleaning, data-driven model discovery and model validation, so that problem owners have more time to devote to solving the business problem at hand.
Applying machine learning to data no longer needs to be an arduous, resource-consuming project spanning several months. The rise of citizen data scientists is bringing significant opportunities for smaller and mid-sized businesses to quickly harness advanced machine learning capabilities to unlock greater insights and business value from their data.