Learning Salesforce Einstein电子书

Title Page

Copyright

Learning Salesforce Einstein

Credits

About the Author

About the Reviewer

www.PacktPub.com

Why subscribe?

Customer Feedback

Preface

What this book covers

What you need for this book

Who this book is for

Conventions

Reader feedback

Customer support

Downloading the example code

Downloading the color images of this book

Errata

Piracy

Questions

Introduction to AI

Artificial Intelligence key terms

Machine Learning

Neural networks

Deep Learning

Natural language processing

Cognitive computing

Pattern recognition

Data mining

GPUs

Programming languages used for machine learning

Practical machine learning with Google Prediction API and Salesforce

Business scenario

Prerequisites

Training and prediction

Integration architecture

Setting authentication for calling API from SFDC

Drawback of this approach

Summary

Role of AI in CRM and Cloud Applications

Sales Cloud Einstein offerings

Automated Activity Capture

Lead Insights

Opportunity Insights

Account Insights

Community Cloud Einstein features

The Company Highlights feature on Chatter

Unanswered questions component for Community Builder

Creating Salesforce Communities

Recommended experts, articles, and topics

Marketing Cloud Einstein features

Social Studio Einstein features

Personalization Builder

Summary

Building Smarter Apps Using PredictionIO and Heroku

Introduction to PredictionIO

PredictionIO platform components

Architecture and integration with applications

Integration with web/mobile applications

Installation of PredictionIO

Prerequisites

Installing and configuring PredictionIO Event Server

Getting started with PredictionIO

PredictionIO DASE components and customization of Engine

Engine design

Query data structure

Predicted response design

Spark MLlib

Data

Algorithm

Serving

Deploying PredictionIO on Heroku

Heroku Buildpack for PredictionIO

Deploying an Event Server application

Deploying the Template Engine

Summary

Product Recommendation Application using PredicitionIO and Salesforce App Cloud

Introduction to Spark MLlib

Setting up the Event Server app on Heroku

Event Server code explanation

Setting up the Recommendation engine application on Heroku

PredictionIO Engine template code explanation

ServerApp

TrainApp

Setting up IntelliJ IDEA IDE for customizing PredictionIO application

Introduction to building Lightning Component for App Cloud and Community Cloud

Visualforce

Lightning Component framework

Component

JavaScript controller

JavaScript Helper

Component CSS file

Apex controller class

Building similar Recommendation Lightning Component for App Cloud

Custom settings for configuration parameters

The ProductViewCapture component

The SimilarProductRecommender component

PredictionIO commands cheat sheet

GitHub references

Summary

Salesforce Einstein Vision

Signing up for Einstein Vision account

Explore Einstein Vision APIs

Creation of dataset

Creating a dataset from a zip file asynchronously

Get status of the upload

Train the dataset

Get status of the training

Prediction with image file

Set up the Heroku add-on for Einstein Vision Services

Authorization setup

Procfile

Obtaining the access token from Private Key

Building Node.js application using Einstein Vision on Heroku using React

Building React UI for image upload

Scaffolding a React App

The index.js file

The App.js file

The results.js file

Middleware using Express

The Episode7 module

The update-token.js file

The fileupload.js file

Testing the application on localhost

Deployment on Heroku instance

Limitations of the application

Summary

Building Applications Using Einstein Vision and Salesforce Force.com Platform

Set up authorization between Salesforce and Einstein Vision APIs

Remote Site settings for Einstein API

Securing Private Key

Apex code utility to obtain access token

Constructing JWT Encoded Body

JWT Bearer token exchange

Creating and training dataset via Apex

Creating dataset using Apex

Monitoring status of training

Train dataset using Apex

Creating an administration app for creating and training dataset

Data model

Application and tabs

Trigger automation for dataset creation and training the model

Creating Lightning Components to recognize image

Summary

Einstein for Analytics Cloud

Setting up Wave Analytics Cloud

Enabling access and permissions to the Analytics Cloud

Creating and assigning permission sets

Creating datasets, lenses, and dashboards

Creating a dataset

Dataflow and data manager

Creating a lens from dataset

Creating interactive dashboards

Scheduling dataflow

Using transformations to create dataset

The sfdcDigest transformation

The sfdcRegister transformation

The append transformation

The augment transformation

The computeExpression transformation

The computeRelative transformation

The delta transformation

The dim2mea transformation

The edgemart transformation

The filter transformation

The flatten transformation

The sliceDataset transformation

An update transformation

Wave Analytics SAQL, XMD 2.0, and dataset Row-Level Security

Salesforce Analytics Query Language

XMD 2.0

Row-level Security for dataset

Introduction to Einstein Data Discovery

Sign up for a trial organization

Importing Salesforce data into Einstein Data Discovery and creating stories

Creating datasets from Salesforce objects

Creating stories

Summary

Einstein and Salesforce IoT Cloud Platform

IoT Cloud key terms

State machine

Orchestration

Traffic view

IoT Cloud components

Input streams and data connections

Data Pipes and data transformation

Orchestrations

Apache Kafka on Heroku

Kafka API

Apache Kafka on Heroku

Supported languages

Node.js sample code for producers and consumers

Encrypting the connection between Kafka and the Heroku web app

Import the Kafka Node.js module

Initializing producer in your Node.js application

Publish interaction events to Kafka

Consuming Kafka messages

IoT integration on the Salesforce Force.com platform

Introducing platform events

Creating platform events

Publish platform events

Subscribe to the platform events

Using CometD to subscribe to platform events

Writing unit Apex tests for platform events

Introducing identity for the Internet of Things

OAuth 2.0 Asset Token Flow for securing connected devices

Prerequisites for implementing asset token flow in Salesforce

Asset token explorer app

OAuth 2.0 authentication flow for applications on limited input devices

Request and Response for device initiating authentication flow

Request and Response samples for polling the token endpoint

Using PredictionIO on IoT events

Summary

Measuring and Testing the Accuracy of Einstein

Measuring the accuracy of Sales Cloud Einstein

Measuring the accuracy of the Einstein Lead Scoring engine

Which lead field values affect conversion rates the most?

Salesforce report to measure the accuracy of Lead Score

Measuring the accuracy of Opportunity Insights

Building evaluation metrics for the PredictionIO systems

ML tuning and evaluation in PredictionIO

Cross Validation

Building the PredictionIO evaluation module

Accuracy

Precision and recall

The f1 score

The confusion matrix

Evaluation in PredictionIO

Measuring the accuracy of Salesforce Einstein Vision

The Get model metrics

The Get model learning curve

Summary