How to build a better bot faster with Dialogflow CX

You may have heard there’s a new kid on the block. At Google Next 2020, Google Cloud have introduced “Dialogflow CX” and the DVELP team have been having some fun experimenting with this new bot-building tool. We’re now ready to share our initial thoughts on what Dialogflow CX could offer our clients.

Dialogflow CX (DF CX) is a new version of Dialogflow, an important tool in the DVELP arsenal. We use Dialogflow to support our clients’ contact centres and help businesses operate more efficiently, improving customer experience and saving costs as a result. (To learn more about how we use Dialogflow, you can watch DVELP’s CEO Tom build a bot live in one of our recent webinars, viewable here).

The old version of Dialogflow has now been separated off and relabeled “Dialogflow ES” (short for Essentials). Google intends that DF CX should be used for building large or complex agents, while ES is better suited for small and simple agents.

The new version is still in beta. Since we already have a vast amount of experience in Dialogflow, we know its features inside out and we were very excited to see what Google’s new features could offer. Now we’ve played around with them a bit, we’ve got a flavour of what we’ll be able to do with it in the future. Here’s our summary:

Dialogflow CX is a state machine-based bot framework. This means it’s much easier to represent the complex trees and paths that power conversational AI. Developers can therefore move faster and implement more features and routes with confidence and reliable results. Moving from the previous context-based flows to the state machine page-based flows is a revolution for Dialogflow bot builders.

With DF CX, we will be able to build the same bots as with ES faster. It offers a smarter and slicker experience at the engineering side. Developers will need less time to deliver a working bot or to make changes, meaning you will be able to iterate and improve your conversational channels at a faster rate.

As well as being a paradigm-shift in terms of how the bot is built, DF CX offers a raft of new features suited for the contact centre, including:

• native DTMF and barge-in support, 
• allowing whole flows to be reused across the bot, and 
• supporting multiple webhook endpoints rather than just the one in Dialogflow ES.

When DF CX is fully up and running, it will likely become our preferred version of Dialogflow. While our clients are all happy using Dialogflow ES, CX is a night-and-day change which both developers and customers will love. 

If natural language processing has got you excited by the potential for improvement in your contact centre, you might like to consider running Intent Capture & Analysis with DVELP. IC&A is a 6 week information-gathering and proof of concept programme in which we help our clients understand customer behaviour and requirements in their contact centre. We are then able to offer low-effort, high-return opportunities for improvement using market-leading AI like Dialogflow from Google Cloud.