✨ PART 3 OF 3 · FLAGSHIP ⏱ 2 MONTHS · 8 WEEKS LIVE AI-NATIVE TRACK

Agentforce
AI .

Build, deploy, and govern AI agents on Salesforce. The future of CRM isn't a chatbot — it's an autonomous agent. Learn to build it.

Duration
2 Months
Modules
9 Core
Level
Advanced
Outcome
AI Builder
// what you'll master

Build production AI agents

Not toy demos. Real agents that ground on enterprise data, respect privacy, and ship.

smart_toy

Build custom agents

Service, Sales, SDR, and fully custom agents from scratch using Agent Builder.

terminal

Prompt engineering

Templates, grounding, RAG, and few-shot patterns that actually reduce hallucination.

hub

Data Cloud + AI

Vector DBs, semantic search, and unified profiles powering personalized agents.

shield

Trust Layer mastery

PII masking, zero data retention, toxicity detection, and audit-ready deployments.

precision_manufacturing

Agent + Flow + Apex

Make agents take real actions — approvals, refunds, callouts via Flow and Apex.

monitoring

Test & monitor

Simulation, analytics, CSAT, and continuous prompt versioning post-deployment.

// curriculum · 9 modules

The Agentforce Roadmap

From Agentforce fundamentals to production-grade deployment & governance.

// week 1 · module_01

Agentforce Foundations

10%
psychology
What is Agentforce?

AI Agents vs Copilots vs Bots, Agentforce Architecture, Use Cases

tips_and_updates Scenario: Explain when to use an Agent vs a Flow vs a Bot for customer support
security
Einstein AI Platform

Einstein Trust Layer, Data Privacy, Responsible AI, Audit Trail

tips_and_updates Scenario: Ensure PII is not sent to LLM during case summarization
groups
Agent Types

Service Agent, Sales Agent, SDR Agent, Custom Agent

tips_and_updates Scenario: Deploy a Service Agent to handle tier-1 support queries automatically
// week 2 · module_02

Agent Builder & Configuration

15%
build
Agent Builder UI

Topics, Actions, Instructions, Testing Console

tips_and_updates Scenario: Configure a custom agent to handle product return requests
topic
Topics & Instructions

Topic Creation, Scope Definition, Natural Language Instructions

tips_and_updates Scenario: Define a topic that handles billing enquiries but not technical issues
flash_on
Agent Actions

Standard Actions, Flow Actions, Apex Actions, Prompt Actions

tips_and_updates Scenario: Add a custom Apex action to fetch real-time inventory data
deployed_code
Channels & Deployment

Messaging In-App & Web, Experience Cloud, Slack, Custom Channels

tips_and_updates Scenario: Deploy an Agentforce agent on a B2C Experience Cloud portal
// week 3 · module_03

Prompt Engineering

12%
edit_note
Prompt Builder Basics

Prompt Templates, Types (Sales Email, Field Generation, Flex)

tips_and_updates Scenario: Create a prompt template to auto-generate a follow-up email after a meeting
database
Grounding & Context

Record Grounding, Related List Grounding, Merge Fields, RAG

tips_and_updates Scenario: Ground a prompt with account and opportunity data before calling the LLM
star
Prompt Best Practices

Instruction Writing, Few-shot Examples, Output Formatting, Hallucination Reduction

tips_and_updates Scenario: Improve prompt accuracy for generating case resolution summaries
api
Prompt Invocation

Apex Invocable, Flow, Lightning Component, API

tips_and_updates Scenario: Invoke a prompt template from a Flow when an opportunity is marked Closed Won
// week 4 · module_04

Conversational AI

10%
chat
Copilot vs Agent

Einstein Copilot Architecture, CRM-native Conversations, Context Window

tips_and_updates Scenario: Use Copilot to summarise a case and suggest the next best action to a rep
forum
Conversation Design

Intent Recognition, Slots, Clarification, Fallback Handling

tips_and_updates Scenario: Design a conversation flow that handles ambiguous customer queries gracefully
memory
Multi-turn Conversations

Session Memory, Context Passing, Handoff to Human

tips_and_updates Scenario: Build an agent that remembers context across a 5-turn support conversation
// week 5 · module_05

Data Cloud & AI

12%
cloud
Data Cloud Basics

Data Streams, Data Model Objects, Unified Profiles, Identity Resolution

tips_and_updates Scenario: Unify customer data from Salesforce CRM and a marketing platform
search
Vector Database & RAG

Vector Embeddings, Semantic Search, Data Cloud Vector DB, Retrieval

tips_and_updates Scenario: Use product knowledge articles as a grounding source for an AI agent
person_pin
AI Grounding with Data Cloud

Connecting Agent to Data Cloud, Querying Unified Profiles, Personalisation

tips_and_updates Scenario: Personalise agent responses using unified customer profile data
// week 6 · module_06

Einstein for Sales & Service

10%
trending_up
Einstein for Sales

Opportunity Scoring, Activity Capture, Sales Emails, Call Summaries

tips_and_updates Scenario: Use Einstein to prioritise top 10 deals likely to close this quarter
support_agent
Einstein for Service

Case Classification, Case Routing, Reply Recommendations, Article Suggestions

tips_and_updates Scenario: Auto-classify incoming cases by type and route to the right queue
lightbulb
Next Best Action

Recommendation Strategies, Action Flows, Display in Lightning

tips_and_updates Scenario: Show a rep the best upsell offer on the account page in real-time
// week 7 · module_07

AI-Powered Automation

10%
account_tree
Flows with AI Actions

Prompt Template Actions in Flow, Decision Logic, Output Handling

tips_and_updates Scenario: Trigger an AI summary generation flow when a case is closed
extension
Agentforce + Flow

Calling Flows from Agent Actions, Passing Variables, Looping

tips_and_updates Scenario: Have an agent trigger a refund approval flow based on conversation outcome
swap_horiz
MuleSoft & Integrations

External System Actions, API Calls from Agent, MuleSoft Anypoint

tips_and_updates Scenario: Allow an agent to check order status from an external ERP via API
// week 8 · module_08

Governance & Testing

11%
shield
Einstein Trust Layer Deep Dive

Data Masking, Zero Data Retention, Toxicity Detection, Audit Logs

tips_and_updates Scenario: Configure trust layer to mask credit card numbers before LLM processing
bug_report
Testing AI Agents

Testing Console, Conversation Simulation, Edge Case Testing, Feedback Loop

tips_and_updates Scenario: Run 50 simulated conversations to validate agent topic coverage
analytics
Monitoring & Optimisation

Agent Analytics, Topic Coverage Reports, Escalation Rate, CSAT

tips_and_updates Scenario: Use agent analytics to identify and fix the top 3 unresolved topics
rocket_launch
Deployment & Change Management

Sandbox to Production, Package Deployment, Version Control for Prompts

tips_and_updates Scenario: Deploy an updated prompt template and agent config without downtime
// why this matters

The market is forming. Right now.

Agentforce is Salesforce's biggest bet of the decade. Enterprises everywhere are scrambling to find engineers who can build, ground, and govern AI agents on the platform — and most teams haven't found anyone yet.

By the end of this 2-month track, you'll be in the top 1% of practitioners worldwide. You don't need permission. You need the skill.

9
Live modules
2026
Up-to-date
100%
Hands-on
1:1
Mentor reviews

Build the agents the world is hiring for.

Two months. Live cohort. Real builds. Get on the inside of the biggest shift Salesforce has ever made.