7 steps to prepare your business for AI

Bernard Marr, the famous strategist and business & technology adviser, describes Artificial Intelligence as “the concept of machines being able to carry out tasks in a way that we could consider ‘smart’”

In the very near future, AI will transform the way we live – whether that’s how our cities operate and flow, how services are delivered, or even the ways in which we interact with buildings and facilities – or them with us. Every activity will be streamlined, connected and optimised through the responsible collection of data and the identification of correlations and potential improvements.

This same logic will then manifest in the way we work. Our productivity, speed, accuracy, insightfulness and innovation will all be enhanced through exploiting patterns in data that would be invisible to the more limited human brain.

But before this ideal can be realised, there are plenty of steps yet to be taken. Each one is important, but none are dramatic or insurmountable. As Marr might say, there is simply more “smart-ness” required.

The truth is that there is plenty more that can derail an AI project before the introduction of technology than after. Here are the seven most critical areas of your business to examine, and the cultural traits required of your business, before you should even consider which technology to use.

  1. Your business strategy

Organizations who are most prepared for AI are those whose strategies encourage the use of data to drive continuous and informed improvements to the business model. Clearly these are the rare breakaway businesses, but the bare minimum requirement will be for organizations to recognize data’s value, and draw occasional insights from it that can offer new directions for the business.

However, despite many businesses’ operations being transactional, predictable and ideal for optimisation and discovering new opportunities, there is often no ambition. This is less a case of insufficient bravery, but more a cultural under-appreciation of the power of data to lead strategy.

  1. Information

Is information a strategic asset in your business, or is it merely a historical record of activities? This is admittedly closely related to the above appreciation of data’s value in determining strategy, but the difference here is the use of information in day-to-day activities. Is it habitually used to improve client engagement or processes? Does its use differentiate you from your competition? Or is information simply use to observe not improve the business?

  1. Heritage in analysis

Almost all businesses analyse their performance and operations, but there are nuances over how. For example, do you analyse data to merely describe what happened in the past, to understand why it happened, or to predict the outcome of future similar activities?

Moving from ad-hoc historical reporting to full AI-driven automation is an enormous leap, and too much for most. But those who already appreciate what is possible with analysis – the art of answering your own questions – will move easiest to AI, which will give you the answers to questions you hadn’t yet thought to ask.

  1. Operational execution

Could you single out a person in your business who is known for being analytical, data-driven and calculated in their approach? If that person is so unusual in your business that they are set apart, then your culture is not yet supportive of AI.

Businesses whose decision makers are well informed with data insights and who will act upon them are already accustomed to trusting the advice of information. Here, AI is probably already in demand in order to speed up the decision-making process. But if your culture is resistant to change, even where the numbers insist upon it, preferring instead to rely on the comfort of the status quo and gut instinct, then this needs to be addressed before investigating AI options.

  1. Architecture

The technical bare minimum for a successful AI deployment is an information architecture framework, ideally one that includes a defined and tested capacity for analytics. The “Vs” of big data (volume, variety, velocity and veracity) are already well-accommodated and both structured and unstructured data can be ingested.

Without this ability to process data, AI is quite simply too much of an ask. Every analyst knows the adage of “rubbish in, rubbish out” when it comes to data flows. It’s no different in AI projects.

  1. Governance

Most businesses are compelled to evidence robust information governance. This will range from local industry-specific regulators, such as in finance, to broader international privacy legislation. But if information governance is largely manual, or if data ownership policies are only piecemeal or out of date, then there can be no confidence in the data’s accuracy or suitability, and therefore nor in the output of any AI project.

  1. Ethics

Right at the beginning of this article, we spoke about how data will influence our day-to-day lives. To many, rather than being seen as valuable, such a concept is disquieting, or even scary.

This is because AI’s influence could be more far-reaching and perhaps less visible than we already predict. This demands the reassurance of a strong ethical framework to surround AI, and a moral structure that is universally agreed upon. These do not yet exist.

These are big ideas, and often considered more esoteric than what applies to a business’ small-scale use of AI. But in fact the same mindset applies. Is appropriate to let AI decide how your customers are treated? Would they expect it? Is the decision-making process transparent enough to control if necessary? Do you have permission to use customers’ data in this way? If you were on the receiving end of the AI project, would you be happy?

Only once you have examined each of these areas will your business be ready to even discuss the most suitable technology. You may be able to hazard an informed guess, but we have all seen projects fail or under-deliver because they were led by assumption and gut instinct. Follow the rigour of the process above, identify your level of maturity and preparedness, and then deploy a technology that you can be certain will succeed.

About the Author

Julian Founded Calligo in January 2012 and is the company’s CEO and CTO. He is responsible for delivering Calligo’s vision of building client-centric business with data services based on the most innovative cloud technologies that optimise the data journey, while ensuring that data privacy continue to be at the heart of the organisation and our services.

He brings over 30 years’ experience helping organisations streamline operations through the innovative application of technology. He has founded several start-up companies, including VirtualizeIT, an award-winning technology company, as well as co-founding Virtustream Inc., a venture capital-backed cloud service provider which raised over $120m in funding from several US-based venture capital institutions and was sold for $1.2B in May 2015.