ChatGPT — Digital Transformation

Image courtesy of metamorworks licensed under Shutterstock

Following on from my previous article ChatGTP — The Next Tech Revolution? I’m going to explore the real world impacts of digital transformation with the direction this tech is taking.

While there are still kwerks with ChatGPT and it often pays to check it’s outputs against a known source of truth. I think it’s fairly safe to assume this tech is here to stay and it’s getting better very rapidly.

With Elon Musk registering another AI company last month (AI.X), it appears he’s re — entering the AI game with the announcement of TruthGTP just days ago on an interview with Tucker Carlson. After his departure from the board of OpenAI some time ago, this is another pivot from Elon. Love him or hate him (I have a healthy respect for him), there’s no disputing that everything he puts his focus on, has a tendency to be seemingly impossibly ambitious. Yet he delivers!

While his key focus appears to be on ensuring AI doesn’t become biased and out of control. This new venture, which will no doubt leverage off his AI work in Tesla cars, robotics and the SpaceX program, will likely help drive the development of lower level commercial and personal use AI capability into another stratosphere.

I have been a heavy user of ChatGPT over the last few months and the improvement seen in the last update is significant. As an example, I used it to write all of the front end public web content for my own SaaS application ProjectSuite. Previously I had to pay copywriters large sums to produce content of that nature. Admittedly it still needs finessing but it’s a pretty good starting point and is especially benefitial for someone like me, who’s generally a lot more comfortable writing code than intelligible sentences!

This brings me to the point of this article. It is obvious this tech is going to be a major player in organisational digital transformation over the next decade and it’s my view that organisations who are exploring their future digital transformation journey (and perhaps more importantly those who aren’t) need to factor this tech in as a major central pillar of their future transformation process.

There are real world opportunities for companies to leverage off this tech, to reduce admin overhead and increase productivity exponentially. An example:

In a call centre environment there are typically frontline staff taking or receiving calls. The actual call generally makes up only one component of the overall interaction with the customer. There would typically be pre and/or post call work required, such as finalising “paperwork” in the companies CRM systems. Apply AI over each call and it could do that off call system work on behalf of the call agent.

Let’s put that into real terms. If an average call handle time is 5 minutes and there’s 1 minute of pre/post call work per call, that’s 20% of time spent in off call work. If AI tech could do that work instead of staff, that leaves 20% more time for the agent to take more calls in a shift. Multiply that across the call volumes over days, months and years, it equates to more customers served, which may equate to more revenue for a sales based company, or alternatively, less call centre staff required to take the same call load, which means less salary costs to perform the function.

Now lets take it another step further. Let’s apply AI over the call in real time and allow it to analyse the speech patterns and tone of the customer, are they positive, negative, what words get the attention of this customer and how do we turn this call around? AI could be sending the call agent real time intel on whether the conversation is likely to result in a sale or not and start giving prompts on what words and phrasing to use that are specific to this customer to help secure a sale.

Better yet, if there is a history with this customer, allow AI access to the history and when a customer calls and identifies themselves through the IVR process (assuming the call centre has this in place). Let it analyze the history to gather intel. Eg. what the customer typically calls for and what the best approach is for this customer. The call can be routed to an agent who's is best suited to deal with this customer and would receive a summary of intelligence on screen before they answer the call.

Again this is manual work an agent would have to do if they need to understand the history of the customer. How many calls to insurance or the bank have you been on where the agent ask you to bear with them while they scroll through and review all of your notes? That takes time, which in the commercial world means more cost, less profit.

All of the above ideas aren’t new. They’re readily available in call centre products right now and, as Rebecca Riserbato points out in her Call Centre AI article, they are identified under the following categories:

1. Predictive Call Routing
2. Interactive Voice Response (IVR)
3. Conversational AI
4. Emotional Intelligence AI
5. AI-Powered Recommendations
6. Call Analytics

A key question Rebecca addresses is; can AI replace call centre agents? In her analysis at the time (mid 2022) Rebecca rightly pointed out that:

“AI can handle repetitive and simple calls by automating a portion, or all of a customer call”.

She continues on to state:

“However, there will always be complicated issues that AI simply can’t handle”.

This statement was absolutely true at the time. However, i’m now shifting my view in line with the progression of this new level of AI. It may very well be able to handle more complex calls than the agents can in the very near future.

My main point is, call centre environments are being revolutionised by the capability of this tech. With serious & focussed investment by the big end of town increasing, we’re likely going to see AI capability grow exponentially in the coming years. We’ll also likely see it explode into every aspect of our daily lives, especially in the commercial environment.

Some parting food for thought… Have a think about the amount of admin overhead that is involved in your daily job or across your company. Now apply an estimated cost for staff to do that. And finally, imagine if a large proportion of it went away tomorrow…