Table of contents

Introduction

Dialogflow CX and ES

2.1.

CX Dialogflow

2.2.

Dialogflow ES

2.3.

Use cases

2.4.

Compare features

2.5.

Pricing

Conversational AI

Agent Assist

Editions

5.1.

Agent types

5.2.

Edition comparison

Frequently Asked Questions

6.1.

What do cloud computing system integrators do?

6.2.

How does on-demand functionality work with cloud computing?

6.3.

How does the term "EUCALYPTUS" apply to cloud computing?

Conclusion

Last Updated: Mar 27, 2024

Dialogflow

Author Shivani Singh

Do you think IIT Guwahati certified course can help you in your career?

Yes

Introduction

Dialogflow is a platform for natural language understanding that makes it simple to develop and combine a communicative UI into your mobile app, web application, device, bot, interactive voice response system, and other applications. You can use Dialogflow to create innovative and inventive ways for members to communicate with your product.

This is a tool that exists outside of GCP and is based on a product called Api.ai. It was created to facilitate human-computer interaction through the use of natural language processing.

Must Recommended Topic, Types of Agents in Artificial Intelligence.

It enables you to create so-called Agents and Intents, which have specific conversation scenarios. Dialogflow can train itself on possible variations of phrases used by the user to demonstrate specific intent. The more phrases are given, the better it will discover to trigger the intent

Dialogflow can analyze various types of customer input, such as text or sound inputs (like from a phone or voice recording). It can also react to your customers in various ways, including text and synthetic speech.

Dialogflow CX and ES

Dialogflow offers two types of virtual agent services, each with its own agent type, user interface, API, client libraries, and documentation:

CX Dialogflow

Dialogflow CX introduces a new approach to agent design, employing a state machine approach. This gives you explicit and clear control over a conversation and a better end-user experience and development workflow.

Summary of notable improvements:

Console visualization: A new visual builder makes creating and maintaining agents easier. Conversation paths are graphed as a state machine model, making them easier to design, improve, and maintain. Conversation states and state transitions are first-class types that provide clear and specific, and powerful control over conversation paths. You can clearly define the sequence of steps you want the end-user to take.
Flows for agent partitions: Flows are first-class types that can be used to replace mega agents. Using flows, you can divide your agent's conversations into smaller chunks. Different team members can own different flows, making creating large and complex agents simple.
Implementation of Omnichannel: Create once, deploy everywhere—in your contact centers and across your digital channels. Integrate your agents seamlessly across platforms such as web, mobile, and messenger and with telephony partners such as Genesys, Avaya, and Cisco.
AI advancements: Enhance your call/chat confinement rate with the newest BERT-based natural language understanding (NLU) models capable of accurately and efficiently recognizing intent and context in more complex use cases.
Complete management: Take care of all your agent management needs, including CI/CD, analytics, experiments, and bot evaluation, within Dialogflow—no other custom software is required.

Dialogflow ES

Google Cloud Dialogflow ES is a complete, build-once, deploy-everywhere development suite for building conversational interfaces for websites, mobile apps, popular messaging platforms, and IoT devices. Dialogflow ES can be used in conjunction with Genesys Cloud to provide customers with new ways to interact with your company by creating engaging natural language experiences through the use of voice-based conversational bots ("Voicebots"). Dialogflow ES incorporates machine learning expertise and products from Google Cloud, such as Google Cloud Speech-to-Text.

Use cases

Customer service voice bots

By developing virtual agents and virtual assistants that can perform tasks such as appointment scheduling, answering common questions, or assisting a customer with simple requests, you can provide customers with 24/7 access to instant communicative self-service, with smooth handoffs to human agents for more complex issues.

B2C conversational chatbots

Connect with your customers on their preferred platform at any time and from any location. Text virtual agents provide an instant and satisfying experience for customers who want quick and accurate responses, whether they want to ask common questions or access specific information.

Compare features

Support multiple languages to connect with your international user/client base
- Dialogflow CX: 30+ supported languages and dialects
- Dialogflow ES: 30+ supported languages and dialects
Analytics: Learn more about how agents perform and how customers interact
- Dialogflow CX: Dashboards for enhanced performance
- Dialogflow ES: Data export to individualized dashboards
Visualizations based on states
- Dialogflow CX: Dashboards for performance
- Dialogflow CX: Data export to individualized dashboards
Build once and deploy across your contact center and digital channels with omnichannel integration
- Dialogflow CX: Integration of various digital channels, such as messenger, messenger apps, and the web. Integration of modern one-click telephony
- Dialogflow ES: Integration with well-known platforms like Twitter, Slack, Google Assistant, etc. Integration of telephony in one click
Prebuilt agents: speed up production time by using a library of agents that have been prepared for common use cases.
- Dialogflow CX: 9 industry-specific production-ready agents for telco, retail, financial services, travel, and other use cases
- Dialogflow ES: 40+ pre-built conversation templates for navigation, hotel booking, etc.
Utilize the top machine learning models created by Google Research for advanced AI
- Dialogflow CX: Natural language understanding (NLU) models based on modern BERT
- Modern models for speech synthesis and recognition
- Dialogflow ES: Natural language understanding (NLU) models of a high standard
- Modern models for speech synthesis and recognition
Visual flow builder: use interactive flow visualizations to easily observe, comprehend, edit, and share work.
- Dialogflow CX: With an easy visual builder for visual state machines, development time can be reduced by 30%.
- Dialogflow ES: bots using forms
State-based models: easily switch subjects and control intricate flows
- Dialogflow CX: Reuse intentions and define transitions and data conditions intuitively
- Dialogflow ES: Simple use scenarios using a flat data model
Additional inquiries: manage conversational detours, then gracefully resume the flow
- Dialogflow CX: Models that can quickly identify and catch tiny conversational digressions
- Dialogflow ES: not authorized
Take care of all your agent organizational requirements inside Dialogflow with end-to-end management.
- Dialogflow CX: Support for flow-level environments and versions during testing and deployment. Native support for tests and traffic splitting
- Dialogflow ES: Basic environments and versions support. Evaluation of virtual agents and experiments is not encouraged.
Speaker Identification: Use biometric voice recognition to recognize and confirm users
- Dialogflow CX: Identify users quickly using only their voice. Boost security with an additional verification layer. By getting rid of passwords and pins, you can reduce user irritation.
- Dialogflow ES: not endorsed

Pricing

Pricing and quotas are described using the terminology below:

Request: Any API call to the Dialogflow service, whether direct with API usage, indirect with integration or console usage, is referred to as a request. The number of queries required for an end-user to complete a job using a Dialogflow agent might vary significantly depending on the task and agent design.
Conversations between an end-user and a Dialogflow agent are referred to as sessions. After the final request for a session is issued, the session is still active and its data is kept for 30 minutes. A session can either be a voice call or a chat session. One chat can be considered as numerous sessions for the purposes of pricing and quota calculations.
Only text is used during a chat conversation for requests and responses. If a chat session receives more than 40 requests, it is divided into numerous sessions, each of which can only receive 40 requests. A session will be considered three sessions, for instance, if it has 81 inquiries.
A voice session uses audio for both requests and responses. A voice session that contains more than 10 minutes of audio is divided into multiple sessions, each lasting no longer than 10 minutes. An audio session with 21 minutes of content, for instance, will be classified as three sessions.
Resource projects and consumer projects: If you use numerous projects, it's possible that the consumer project linked to your request authentication won't be the same project linked to the request's agent (resource project). In this instance, quotas and prices are established using the consumer project.

Conversational AI

Dialogflow is a component of Google Cloud's Conversational AI offering.

Google Cloud takes four approaches to Conversational AI:

Customer service: Chatbots, voice bots, and telephony helpdesks, built into Contact Center AI offerings such as Dialogflow, Agent Assist, and CCAI Insights, can help you resolve your customers' needs faster.
Internet of Things and custom hardware: To simplify integration, add voice components to your car navigation systems, kiosk environments, smartwatches, and more using our APIs, which include Speech-To-Text, Text-To-Speech, NLP, and Dialogflow SDK.
Engines of discovery: Conversational AI expands the possibilities of search channels, enabling opportunities such as Conversational Ads on Google Search, Google Maps, and Business Messages.
Assistive voice: Allow your customers to do anything from adding tasks to their to-do list to having turned on lights.

Agent Assist

Virtual agent services are provided by Dialogflow CX and ES for automation and contact centers. If you have a contact center with human agents, you can use Agent Assist to assist them. While human agents are conversing with end-user customers, Agent Assist provides real-time suggestions.

The Agent Assist API is built on top of the Dialogflow ES API. These additional types and methods will be visible when browsing the Dialogflow ES API. Despite the fact that Agent Assist is an extension of the Dialogflow ES API, a Dialogflow CX agent type can be used as the virtual agent for Agent Assist. You can ignore these extensions if you are only using a Dialogflow virtual agent.

Editions

There are various agent kinds and editions of Dialogflow agents available. Each version has different agent types, features, prices, and quotas.

Agent types

There are various agent kinds and editions of Dialogflow agents available. Each version has different agent types, features, prices, and quotas.

There are several different sorts of agents:

Agent CX

This kind of sophisticated agent is appropriate for big or extremely complicated agents. The building elements of conversation design are flows and pages, and state handlers are employed to regulate discussion pathways. The Dialogflow CX basics provide an overview of the CX agent type. Some of the features of agent CX are the following:

Its intention structure is flat.
Most of the console's controls are text-based.
Here, the intentions are linked to actions, events, and reactions; they are particular to the state of the discourse, making them impossible to reuse.
When errors occur, the agent quietly ignores them and passes them along to the API caller if they are large enough.
The parameter scope's ability to be applied to intent, context, or events makes it flexible.
Here webhook calls are necessary.
It has granular Pricing and quotas.
It has moderately complex agents.

Agent ES

This is the common agent class, which works for small to medium-sized, straightforward to moderately complex agents. Contexts are employed to steer conversational trajectories, while intentions serve as the cornerstones of conversation design. The Dialogflow ES basics provide a summary of the ES agent type. Some of the features of agent CX are the following:

The console uses visual graphs depicting communication channels and text forms for configurations.
It includes a graph structure of flows and pages.
If your agent has explicit error event handling built-in, the intents are simplified to eliminate this coupling and made to be extremely reusable.
The scope of the parameter allows it to be applied to intent, forms, or sessions.
It can be set up statically in fulfillment, with static route conditions, or through webhooks calls
It has rendered pricing and quotas simpler.
It has complex agents.

Edition comparison

There are the next editions listed below:

The trial version of Dialogflow: A free version with the majority of the features included in the common ES agent type. It provides a constrained quota as well as community and email support. Dialogflow testing can be done with this edition.

For Dialogflow ES: The pay-as-you-go Dialogflow Essentials (ES) Edition offers the typical ES agent type. Production-ready quotas and support from Google Cloud have been included in the Essentials Edition.

ConversationFlow CX Edition: The sophisticated CX agent type is offered by the pay-as-you-go Dialogflow Customer Experience (CX) Edition. Production-ready quotas and Google Cloud support are available with the CX Edition.

Frequently Asked Questions

What do cloud computing system integrators do?

The cloud may be made up of numerous complicated components. The cloud's system integrator is a method that, among other things, enables the process of developing the cloud and integrating its many components to produce a hybrid or private cloud network.

How does on-demand functionality work with cloud computing?

Cloud computing was developed as a technology with the goal of giving all of its user's access to functionality whenever and wherever they need it. It has accomplished that goal thanks to recent developments and the simplicity with which tools like Google Cloud are accessible.

How does the term "EUCALYPTUS" apply to cloud computing?

Cloud clusters are deployed using the open-source cloud computing architecture known as "EUCALYPTUS," which stands for "Elastic Utility Computing Architecture for Linking Your Programs To Useful Systems". You can create public, private, and hybrid cloud systems with the "EUCALYPTUS." You can even set up your own data center on the cloud, which you can employ to take advantage of its capabilities for your business.

Conclusion

To sum it up, in this blog we discussed the basics of Dialogflow, and its two agent services- Dialogflow CX and ES. We also saw the properties of Dialogflow CX and ES. Then we discussed conversational AI, agent assist, and last but not least editions.

For more content, Refer to our guided paths on Coding Ninjas Studio to upskill yourself.

Do upvote our blogs if you find them helpful and engaging!

Happy Learning!