Last month we looked at the role of RPA in helping organisations come out of their Covid-19 isolation, and in this post we want to focus on the same but for Artificial Intelligence (AI). At the time of writing, there are various degrees of the lockdown restrictions being lifted across Europe, with those countries that were impacted first, such as Italy, returning to some sort of normality, whilst others, such as the UK, only just starting to take those initial steps. But what everyone has now started to do is look at their ‘exit strategy’ – this is not just a case of understanding how can we get back to some sort of normality as easily as possible, but looking at how some of the temporary-but-transformational measures that were put in place could potentially be made permanent.
It’s clear that a lot can be achieved when there is a crisis that threatens the survival of businesses: everyone collaborates, pulls together and gets things done in days rather than weeks, and in weeks rather than months. Many people would have thought that that sort of accelerated implementation wouldn’t have been possible with AI. Normally AI projects take months to prepare and execute – collating data, building the models, fine-tuning, testing, etc is inherently a pain-staking effort. But some AI can be done relatively quickly, especially if a lot of that design and build work has already been done by a vendor.
As an example, a health charity was facing a huge crisis as its donations had dried up overnight due to the uncertainty that the pandemic had brought. New funds were required quickly, but in this sort of situation you only get one real chance to ask for the money, so you need to make sure you are asking the right people and using the right messaging. The charity brought in a software vendor who had already built algorithms that were able to predict people’s propensity to donate. All the vendor needed was the data from the charity’s own CRM system to make those predictions for their donors. From the first discussion to the models being built ended up being less than a month.
The charity also needed to ensure that the messaging was aligned with the issues and concerns of their community. During the pandemic, the situation was changing week by week, so it was really tricky to identify the really important stuff. By using ‘comparative linguistics’ from a specialised AI vendor, they were able to monitor their forums and social media feeds to identify what topics and themes were becoming less or more important each week. Issues such as access to medical professionals was important early on, but moved onto the more basic requirement of the ability to buy food. Most recently the core themes have been around insecurity and stress. By understanding this, the charity was able to align their messaging with their community’s so that they can provide appropriate support and target their donation requests sensitively.
That ability to do things quickly and simply shouldn’t stop once the lockdown is eased or over. AI can be instrumental in enabling your business to cope with the potential wave of ‘routine’ work that had been put on hold during lockdown that will now come flooding in. The ability to triage this work – prioritising it and channeling it to the right resources to deal with – will enable you to focus on the right things. AI can, in a relatively short time, learn to recognise and classify different types of documents coming into the contact centre, and can analyse emails for sentiment and intent. Automating some of the downstream work after triage, using RPA for example, will increase efficiencies even further.
Some applications of AI will, of course, still be tricky to do and will take time and effort. But there are many that do not. Think about how these could help you now, but also if there is another wave of coronavirus, how they could help you cope better the second time around. Building chatbots, automating processes, and predicting demand, for example, will put you in a perfect position should we find ourselves in a similar situation. And, don’t forget that it will make your business more efficient and better able to serve your customers in the meantime.