We are all customers and we have all had a problem at some stage that needed expert help. Most of us shy away from making that call we know we have to make. It is only the brave and the patient that embark on the maze of the IVR (interactive voice response) system in an attempt to speak to a human being that can solve the problem.

We, as industry professionals, are all aware of the initiatives, surveys and (let us be honest) hype around omnichannel care – the concept that customers should be supported via any channel they wish, whether it be Twitter, Facebook, phone, text or even instant message.

We also know that early attempts to automate this process have not been completely successful. High profile failures in the implementation of bots to support customers and customer service operations have been widely published.

This, of course, raises the question – what can be done about it?

The answer could well lie in a key TM Forum project known as the Cognitive Digital Agent Catalyst.

The goal of this Catalyst is definitely a challenging one. Especially in view of its vision statement which outlines the aim of "creating the next generation of customer care service through artificial intelligence".

The team behind this Catalyst (championed by Globe Telecom and Verizon and supported by industry participants Huawei, Amdocs and Infosys) has taken a thoroughly practical approach to the problem. They have taken the customers’ challenges – endless IVR menus and hard to navigate websites – and matched them with the ‘mirror’ challenges faced by service providers – how to drive more customers to self-care channels, and driving a lower ‘cost to serve’.

By doing so, they have created a very compelling project, that will create a win-win environment.

Essentially, the goal is to "explore the use of cognitive computing or artificial intelligence technologies such as natural language processing and machine learning to create the next generation of customer care services".

One of the key parts of this initiative is the creation of a cognitive digital agent that can answer questions on various channels, including IVR, and only funnel more complex queries or those that cannot be resolved automatically to human beings. It therefore needs to understand natural human voice and use artificial intelligence, machine learning and the latest CRM techniques to produce the results.

Training a digital agent

As such, a major focus will be on the machine training aspects, i.e. what are the best ways to train a cognitive digital agent so that it can genuinely support and improve the customer experience? This is not only critical to the project but the most difficult aspect to address.

The final result aims to improve the customer experience greatly, and even make it fun to interact through the channels that used to be so irritating. It will also be an excellent test of how artificial intelligence and machine learning can augment – rather than replace – the human workforce.

In practical terms, the team is considering how to use the TM Forum’s Open APIs as integration tools for the various elements that are needed. They will also draw on the omnichannel work that has already been carried out by previous Catalysts. The single cognitive computing platform, linked to the CRM will not only work towards breaking down silos but also prove that implementation timescales can be dramatically reduced.

This Catalyst could well evolve into something that becomes key to many areas where common problems exist within service providers. The best practices and learnings that will be part of the deliverables will be fed back to TM Forum and used in other initiatives.

This is an area that is not being explored extensively enough within the industry, and hopefully the next report will announce that mass adoption of cognitive computing technologies is a reality.

The Cognitive Digital Agent Catalyst will be demonstrated at TM Forum Live! Asia in Singapore next week.

For more information, watch this video interview: