In February 2020, Google announced the launch of Dialogflow Mega Agent in beta for developers and businesses. Dialogflow is Google’s AI- and ML-driven technology for building natural conversation experiences. It allows organisations to build powerful chat- and voice bots they can deploy on websites, mobile apps, the Google Assistant, Facebook Messenger and other popular platforms and devices. But what is this brand new Mega Agent feature in Dialogflow about, and how can it help Contact Centers? I’ll explain in this article.
Intents in Dialogflow
First of all, let’s take a look at what an intent is in Dialogflow, because the added value of Mega Agent lies in these intents.
An intent categorises an end-user’s intention for one conversation turn. For each Dialogflow agent, a virtual agent that understands natural language and handles conversations with your end-users, you define many intents. Your combined intents can handle a complete conversation. When an end-user writes or says something, referred to as an end-user expression, Dialogflow matches the end-user expression to the best intent in your agent. – Dialogflow Documentation
If you are a bit familiar with Dialogflow you already know that the platform originally limits the number of intents to 2,000. But with the introduction of this Mega Agent feature, the number of usable intents can increase up to 20,000.
Contact Center and Customer Support Chatbots
This is a huge change of paradigm for Chatbots that are supposed to handle a very large number of intents. Among all the possible use cases, Contact Center and Customer support chatbots will definitely be the most impacted. Why? Because this kind of chatbots can easily exceed the maximum of 2,000 allowed intents in the Dialogflow standard edition.
With the introduction of this feature in beta, Google Contact Center AI has the possibility to build much more complex and intelligent Dialogflow bots, and make them able to handle a much more important amount of intents.
Google Contact Center AI
But what is Google Contact Center AI? Contact Center AI is a feature powered by Google Cloud, released at the end of 2019. Designed to help contact centers with artificial intelligence technology, this AI-based tool is engined by Google Dialogflow for the conversational layer.
Contact Center AI aims to help companies in the daily basic interactions they have with their clients. It acts as a virtual “Customer support agent” that offers AI-based conversations as a service by Google Cloud. Contact Center AI can understand what customers say with speech-to-text, respond naturally with text-to-Speech and ultimately take action with Dialogflow.
Dialogflow Mega Agent & how it works
So how does it work, the new Mega Agent functionality in Dialogflow? You can combine multiple Dialogflow agents, called Sub agents, into a single agent, called a Mega Agent. When a user interacts with the Mega agent bot, the bot will perform a detect intent request against the Mega Agent, alongside all of the Sub agents.
And this is where the magic happens: the fact that all the Sub agents are also being considered. When this request happens, the best possible response from all the considered Sub agents will be returned to the Mega agent.
Impact of Mega Agent for Customer Support
The biggest impact of this feature is, in my opinion, the fact that several teams can manage their own particular knowledge area within a Mega agent. It means that each team can manage a subject, use case or topic covered by the scope of the bot. Imagine a setup where you create one Sub agent per subject/use case/area (ex: HR, Billing, PR, Accounting, Marketing, etc.) and give management responsibility to a person within each team to ensure better governance.
Then you aggregate all these Sub agents, into one single Mega Agent. This seems like the perfect way to manage complex customer support, isn’t it?
Advantages of using Dialogflow Mega Agents
There are a two main reasons that you may want to use Mega agents:
- Better governance: as previously said, if you have multiple teams building and/or managing a bot, each team can be responsible for one Sub agent. This helps to improve governance as chatbots with large numbers of intents are difficult to maintain. However, with the Mega agent feature, each team can be responsible for one sub-agent, which will simplify change conflicts across teams and limit the impact of any potential issue to the initial Sub agent. In simple words, it increases bot management efficiency and mitigates risks to break the bot.
- More intents: If you have a bot with a large number of intents, there is an important risk that you may approach the intent count limit (2,000 intents). In this case, creating multiple Sub agents dependent on one Mega agent might be the best solution for you as your intent limit of 2,000 is multiplied by the amount of Sub agents you are using.
However, the huge advantages of Mega agents are balanced with some limitations to consider when scoping the full architecture of the bot :
- As with all agents, you can only create one agent per Google Cloud Platform (GCP) project, so each Sub Agent and Mega Agent will be associated with its own GCP project.
- A Mega Agent can have at most 10 Sub agents. It means that the total of intents that can be handled with one Mega Agent linked to 10 Sub agents is 20,000 intents.
- The Small talk feature which is usable by default for each Dialogflow agent doesn’t work for Mega Agents.
Empowering your Contact Center Employees
For some time now, Google has been supporting Call center employees. The Contact Center AI solution is definitely the best service companies can use to increase operational efficiency and personalised customer support.
The release of the Dialogflow Mega Agent feature is another sign of Google’s desire to leverage even more Customer support experience for both employees and clients. Just because customer support should never provide a bad user experience.
Better governance. Risks mitigation. Quality improvement. The advantages of the Dialogflow Mega Agent feature released in beta a few months ago are already pretty impressive. And also perfectly tailored for Contact Center AI, which is one of the most promising features of Google Cloud, released for all countries at the end of 2019. These quick wins brought by Dialogflow Mega Agents, coupled with the Contact Center AI service offered by Google Cloud will improve your customer support and make the lives of your Customer support employees much easier.
Do you want to set up your own FAQ chatbot as a rapid response bot for COVID-19 related questions?
Any questions about Dialogflow or do you need guidance to set up your own chat-or voicebot?
We’re happy to help!