How real companies are using AI every single day

You might think AI is reserved for the big fancy companies like Amazon or Apple or Facebook but actually, AI is used by hundreds of companies, big and small, in the present day with millions of interactions with customers – but how can we use this amazing technology to make everyday life for our customers and employees better?  

There’s lots of use cases for AI, one being the biggest time saver as an AI analysis of billions of different molecules, chemically binding to the target protein to create COVID vaccines – humans couldn’t possibly do that in the same timeframe!  

Another example is how Netflix analyses everything you watch, when you turn it on, when you turn off, when you pause, when you rewind and fast forward, how long you watch the show for, what kind of show, the genre, the mood, the actors. It uses that information to give hyper personalised recommendations of what it thinks you’ll like. And they are so good at it that 80% of the views come from its AI recommendations. But did you know that everyone gets different thumbnails?  

Netflix uses an elaborate AI thumbnail selection process for each of its 200m+ users. The process is called Aesthetic Visual Analysis or AVA. AVA is a set of tools and algorithms that search Netflix videos for the best images which are graded on three variables: 

• Visual - which is brightness, contrast, colour, motion blur 
• Contextual - which is face detection / shot angle 
• Composition - which is photography principles like "rule of thirds", symmetry, depth of field 

The next step is "Image Ranking", which chooses the thumbnails most likely to be clicked. These are going to be the ones with expressive faces and main characters which is all very fancy and advanced. But even at the most basic level, Netflix uses Machine Learning to choose a thumbnail based on your watch history. 

It’s worth noting that AI is not just for the mega brands but something that can be utilised by nearly every business to materially improve their customers experiences and make life easier for their employees. Let’s talk about one of the simplest routes of communication; emails.  

The speed of which most businesses are able to respond is frankly embarrassing – it’s far too slow. We know most companies expect the email channel to fade away in time but it’s certainly not going anywhere right now. And since the pandemic, customer’s patience is at an all-time low according to a recent Pega study on real-time decisioning, so we have to find ways of speeding up the resolution times.  

The problem in my opinion is most organisations struggle with the unstructured nature of email and the manual work needed to manage the workflow. In my experience most email work sits siloed outside the system of record and the companies standard processing models causing a lack of visibility and bottlenecks 

To solve the problem, some companies are turning to AI to triage, automate and speed up the end-to-end process. They are using it to eliminate the lengthy delays customers often experience by identifying the actual intent and sentiment of the email automatically. Imagine that – AI is reading the content of the email, assigning work and pushing to the next stage in the process. Just imagine how much time and effort that would save. 

That is exactly what one of the largest financial institutions in Australia decided to do. National Australia Bank had embarked on a multi-year transformation program focused on making itself simpler and faster with a big focus on shifting manual human driven work to digital process automation. They very quickly saw the advantage of deploying these email bots to reduce the backlog, speed up customer resolution and eliminate lost and duplicated requests – and they saw some fantastic results incredibly quickly. 

Key Benefits: 

• Triage of thousands of emails per day (~120k per month). 
• And saw an astonishing 80% reduction in manual triage. 
• And this wasn’t just a one size fits all – they deployed: 
  • 61 active email bots serving hundreds of mailboxes across 500+ case types 
  • Enabling 4,500 concurrent users across 22 business units. 

As a glorified connect-the-dots game, the way AI here analyses the masses of emails so quickly is similar in fashion to how the Netflix AI analyses the frames and thumbnails. In both scenarios, this has already had a significant impact on the customer satisfaction and the need for agent intervention has radically reduced. So, there you have it – that’s how real companies are using AI every single day to make their employees and customer lives easier and save time and money along the way.