In the last few weeks, we've worked with a number of companies that built and deployed their own natural language IVR or chatbot using a cloud based platform such as Google Dialogueflow, Amazon Lex or Twilio Autopilot and now have either hit the limitations or have realized how hard it is to create quality Customer Experiences without hiring a multifaceted team to manage the entire workflow.

While the Do-It-Yourself bot solutions many seem like a good way to triage a call center during times of high call volume or needing to reduce agent costs, most companies hit the wall with this approach pretty quickly.

Some comments we regularly hear from people that have tried this approach:

  • I couldn't get it to recognize a spoken address
  • You have to feed it a lot training data for it catch the intent. We don't have that much data, and don't have the resources to continue to train it.
  • It can't handle alphanumerics, such as an account number or order number
  • I can't even tell when its working or what customers are experiencing
  • It seems cool, but not ready for prime time yet
  • It's clearly not the same technology that powers Alexa or Google Home.
  • There's no reporting, so my operations manager killed the project
  • The data ownership and use is unclear. We're not sure what Amazon, Twilio or Google do with the data from the calls or chats
  • We gave up trying to make it work

For companies that need to build a voice application which goes beyond a simple demo and requires models other than generic, pre-trained consumer domains, cloud-based NLP services are typically not the best approach.

These are pains we've felt ourselves. What most don't realize, is that Xaqt actually started our Conversational AI journey using Amazon Lex, Google DialoqueFlow, Twilio Autopilot and Rasa. We had committed to building a Conversational IVR for the City of Kansas City, MO's 311 call center, but we couldn't get any of the do-it-yourself products to provide a satisfactory customer experience. Plus, they're expensive to operate.

That's why we built our own. And it's why we set out to create a framework with industry leading accuracy and ease of use.

Your customers deserve the best experience possible, and your internal software engineering resources probably have limited bandwidth to learn how to train, maintain and build all of the ancillary tools needed in order to provide that quality customer experience.

And that's why we're launching our new bot conversion tool and services.