Machine Intelligence in Contact Centers

Machine Intelligence in Contact Centers

by Evan Macmillan – Gridspace

“The Internet is ultimately about innovation and integration, but you don’t get the innovation unless you integrate Web technology into the processes by which you run your business.” 
— Louis Vincent Gerstner Jr.

A former longstanding CEO with American Express and IBM, Louis Vincent Gerstner Jr recognised the link between integration and innovation before many twenty-first century executives did. The best companies would not only use the Internet; they would integrate it into the cores of their organisations. This would allow them to harness Internet-age technology to reinvigorate stale processes and benefit their customers in new ways.

These days, increasing numbers of companies are coming around to Gerstner’s thinking with respect to machine intelligence; however, they are continue to struggle with the integration element. In a recent survey of global contact centres, nearly half cited integration as their greatest challenge.

But what does integration mean in this case? The integration challenge facing contact centres seems more daunting: AI services, across greater numbers of channels, architectures and regions. Even customers themselves seem more daunting, as well. Today’s customers consider the AI systems from Google and Apple as benchmarks for the IVR systems (voice dialogue systems) of mutual savings banks.

A few years ago, when we first started thinking about machine intelligence in contact centres, we asked companies their thoughts on what made the integration of intelligence so challenging. Why weren’t intelligent services, such as real-time conversational ASR and natural language extraction, better suited to integration? What we learned was fascinating, and it influenced how we developed our product, Gridspace Sift.

Following on from this, we looked into how the average contact centre approached technology integration. It transpired that most customer contact centres focussed on static customer journeys and outcomes. This was why, for instance, two years was appropriate for the training and rolling out a natural language interactive voice dialogue system. Furthermore, we learned how contact centres evaluated new integrations (usually by hand) and how those integrations improved with time (they didn’t always).

Finally, we learned about the people – both the on-site integrators and contact centre operators. We met individuals with an incredible amount of institutional knowledge about their customers; however, we were surprised by how difficult it was for them to apply this knowledge and to influence customer-facing processes.

Because contact centre integration was so hard, we observed a disconnection between great ideas and implementation. We saw people integrating and operating the wrong things really well. To quote Tom and David Kelley’s warning on two-year prototypes, “The more you invest…and the closer to ‘completion’ it is, the harder it becomes to drop an idea that’s not working.” We witnessed an industry that was poised for innovation, but inhibited by integration.

If you still haven’t tried Gridspace Sift, we believe you will love it. Here’s a quick concept to get you started, and to demonstrate how simple it is to build conversationally aware customer service interactions. This script allows you to find pricing quotations in real-time using speech. You try it out at – and it won’t take two years to test :)

After uploading and calling up the script, simply follow the prompts.

Add a comment

your browser is not up to date
to enjoy this website you will need to install a modern browser.
we recommend to update your browser and to install the latest version.

iOS users, please male sure you're running at least iOS 9.

Mozilla Firefox Google Chrome Microsoft Edge Internet Explorer