Echo Dot (3rd Gen) - Smart speaker with Alexa - Charcoal

Use your voice to play a song, artist, or genre through Amazon Music, Apple Music, Spotify, Pandora, and others. With compatible Echo devices in different rooms, you can fill your whole home with music.

Buy Now

Wireless Rechargeable Battery Powered WiFi Camera.

Wireless Rechargeable Battery Powered WiFi Camera is home security camera system lets you listen in and talk back through the built in speaker and microphone that work directly through your iPhone or Android Mic.

Buy Now

Enterprise Virtual Agents with Google DialogFlow Mega Agent

0
129


Bharat S Raj

With the advent of AI, Enterprise Service Desk/Contact centers are exploring how they can provide great customer interactions with the increasing number of challenges they face. Chatbots/Virtual Agents are one such solution that provides great UX at an optimal investment.

On 19th February 2020, Google announced the launch of Dialogflow Mega Agent in beta for developers and businesses. In this article, let’s explore more in detail about Google DialogFlow Mega Agent and how your organization can use it.

For beginners, Dialogflow is a natural language processing (NLP) platform that can be used to build conversational applications like chatbots, virtual agents in various languages and on multiple platforms. The company was previously called API.AI and was acquired by Google in 2016. It was then renamed Dialogflow.

The Dialogflow platform originally limits the number of intents to 2000, but with Dialogflow Mega Agent you can increase 10 fold to 20,000 intents. This is a significant development for chatbots which have a large number of intents > 2000. This means that a business using Contact Center AI can make a complex Dialogflow Mega Agent and use it across a number of use cases, or multiple agents can be combined into a single agent. E.g. A Enterprise can have multiple bots incorporated into a single bot.

Sample Mega Agent with multiple Sub Agent(s)

This means that different teams can manage a particular knowledge area for one chatbot subject, use-case or topic area. You can combine multiple Dialogflow agents, called sub-agents, into a single agent, called a mega agent. When the users type a question to the chatbot, all of the sub-agents intent are considered, and the best response from the sub-agent is returned.

This also helps in better governance — chatbots with a large number of intents are difficult to maintain and it’s dependent on the development team to make the fixes. However, with DialogFlow Mega Agent each team can be responsible for one sub-agent, which simplifies change conflicts across teams.

Why 2000+ intents is a big deal? Why not use Knowledge Connectors

Most Enterprise Chatbot projects have a shared need — the ability of the bot to answer a large number of questions. These questions/answers are often available in FAQ pages or PDF documents or any other unstructured data sources. Technically, the answers are available in the knowledge base however bots have a challenge of understanding these unstructured data sources. The challenge for a successful chatbot is utilizing this often unstructured information to understand a question and provide the correct answer.

Knowledge connectors are a beta feature released in 2019 that allows various documents sources to be added to your agent — they parse knowledge documents to find automated responses. The knowledge data source(s) can be a document (currently supported content types are text/CSV, text/HTML, application/pdf, plain text) or a web URL which has been provided to the Dialogflow agent. Knowledge connectors can be configured either through the web console or using the client library that is available in Java, node.js & python

The limitation with Knowledge connectors is that it offers less response precision and control than intents. There can be more than one way of asking a question but the answer can be the same. Extractive QA may not work always and results are disappointing — the extracted answers look more like a match based on keywords with some additional coverage but it does not appear to consider the context in which the question is asked. Knowledge connectors are an experimental feature, so hopefully as the technology advances then they will improve.

With 2000+ intents, you can supplement your bot’s ability to answer more questions and actions. You can bring in your own set of logics for knowledge extract and build a smarter bot for your use case.

Getting Started with Mega Agent

Create a new agent and mark it as mega-agent.

Then, click on the Sub-agents link under the agent name. Once the agent is selected then a Sub Agent button is enabled. Click on that and add all your sub-agents.

When choosing adding sub-agents you can select an environment or whether to include or exclude the knowledge Base.

This is truly a gamechanger for developing Enterprise Chatbots as now they can build better chatbots with 2000+ intents. In most enterprises, chatbots are limited to a particular department or functionality or use case. This feature allows multiple agents to incorporate into a single agent.



Read More

LEAVE A REPLY

Please enter your comment!
Please enter your name here