Skip to main content
Gruppe

Agentblazer Community Group

Welcome, Agentblazer! Are you ready to connect with AI leaders and enthusiasts from around the world? You've come to the right place. The Agentblazer Community is the premier destination to skill up on AI, discuss best practices, gain access to Agentforce product experts, showcase your expertise, and supercharge your career.

Who owns it when an AI agent gets it wrong? (A question I keep coming back to.) 

 

We spend a lot of energy on the happy path — the agent qualifies the lead, drafts the reply, updates the record. Far less on the moment it does all of that confidently and incorrectly, maybe in front of a customer. 

 

Agents act in seconds. The human judgment to catch and override them scales in months. That gap seems to be where a lot of rollouts quietly fall apart. 

 

I'm curious how teams here are handling it in practice: 

 

1. Who actually owns or reviews an agent's decisions in your org — RevOps, the line manager, a dedicated "AI steward", or nobody yet? 

2. How fast can a human step in when an agent goes off the rails — and is there a clean rollback? 

 

Not after the vendor answer — genuinely want to hear how Salesforce teams are governing this on the ground. What's working, and what isn't? 

 

#Agentforce #Salesforceadmins

 

 

1 Kommentar
  1. 19. Juni, 12:18

     We are rolling out a new Agent that, in testing, has a 93% success rate in routing emails. These are results than expected for the introduction but the busines says, what about the 7% that are "wrong".  

     

    In our partnership with Salesforce they asked/told us early in the project, you will need to have an "Agent Manager". This person or persons would be responsible for ensuring the agent stays within it's guardrails, monitors feedback from users etc. etc. etc. . Not quite a baby sitter but something close to it. So who should be in that role, someone from the business, an Admin or Developer??? We haven't identified an Agent Manager.... yet.

0/9000

The Agentblazer Collection 🗞️

     

What is it? 

The Agentblazer Collection is a weekly post shared in the community with resources from Salesforce related to Agentforce, community-contributed resources, and other exciting items ✨   

   

Community Contributed Resources:

For a resource be considered for a future post of The Agentblazer Collection, please share content with us by using the workflow

 

Featured Resources this Week:

 

Hope your week is going well!

0/9000

AI agents went mainstream in 2026. The ROI didn't follow. What are you actually seeing? 

 

The data this year is hard to ignore: ~97% of companies deployed AI agents, but only ~29% see real ROI from generative AI (23% from agents), and Gartner now expects 40%+ of agentic AI projects to be cancelled by 2027. 

 

In my experience the failures rarely come down to the model. They come down to three things underneath it: 

 

1. Data nobody trusts. You can't automate decisions on a pipeline your own team second-guesses. A lot of teams are realising their CRM data just isn't clean enough to feed an agent yet. 

 

2. Strategy that's "for show." 75% of execs admit their AI strategy is more performance than guidance. Agents pointed at fuzzy goals fail fast. 

 

3. No clear owner for "when is the agent wrong?" Agents act in seconds; the judgment to override them scales in months. 

 

The teams getting real value seem to do the boring things first: trusted data, one scoped pilot with one metric, and a named human who owns the override. 

 

Two questions for the community: 

 

1. If an agent rollout stalled for you, what was the real reason — data, strategy, or ownership? 

2. How is your team handling the "who decides the agent is wrong?" question? 

 

#Agentforce #Salesforce

1 Kommentar
  1. Gestern 06:26

    Hi @Miteshkumar V Jain

     

    After reading all the numbers and stats, i got a mixed up reaction. The facts mentioned by you seem to be alinged toward a specific outcome, though actual report says the same numbers in a different way. 

    Here are the actual statements (Section "Summarized by WRITER" Source =

    https://writer.com/blog/enterprise-ai-adoption-2026/

    :  

     

    • 97% of executives deployed AI agents in the past year, with 52% of employees already using them.
    • Only 29% see significant ROI from generative AI, despite individual productivity gains of 5X

    52% already using them, and 5x productivity gains seem to be ommited. Also, the following statement about Gartner is from an year old Gartner report dated June 2025 (source https://www.gartner.com/en/newsroom/press-releases/2025-06-25-gartner-predicts-over-40-percent-of-agentic-ai-projects-will-be-canceled-by-end-of-2027), and seem outdate now after launch and evolution of latest Agentic platforms and AI Models. 

    In our experience, we saw few agent rollout stalled almost one year ago (due to lack in confidence on those AI automations), but since then we have implmented more than 10 succesfull implementations, that are running live in production now. My belief is, if you are well-aware and confident of what you want to do with AI agents, then it has higher probability of achieving its targets.  

0/9000

Introducing Agentforce Voice: AI-Powered Conversations at Scale 📞 

 

Voice is the next frontier for customer experience, and we’re bringing intelligent, natural voice capabilities directly into your workflows. Learn how to configure seamless, AI-powered voice agents that can handle complex inquiries and scale your team's capacity without losing the human touch.  

 

Discover exactly how you can easily build, test, and manage these voice interactions natively in Salesforce to drive faster, smarter resolutions. 

 

👉 Read the blog post

0/9000

Hi everyone,

Can we create Salesforce CPQ Quotes using Agentforce? In other words, can an Agentforce agent access and perform CRUD operations on CPQ objects such as SBQQ__Quote__c?

I tried granting object-level permissions on the Quote object to an Agentforce user, but I'm getting the following error:

"Can't save permission set... The user license doesn't allow the permission: Read/Create/Edit Quotes. This permission set contains custom object(s) 'Quote' that require a license."

The user already has the Salesforce CPQ License assigned. Has anyone successfully integrated Agentforce with CPQ objects or encountered this limitation? Is this a licensing restriction of the Einstein Agent user license, or is there additional configuration required?

Thanks in advance! 

Can we create Salesforce CPQ Quotes using Agentforce?

 

 

1 Antwort
0/9000

Summer '26 is live. What's your week-1 plan? 

 

The temptation is to turn on everything at once. After enough release cycles, I've learned the orgs that get value early move in a deliberate order instead. Here's the sequence I'd run: 

 

1. Audit custom Apex for the API v67.0 change first. SOQL/SOSL/DML now default to user mode, and Apex classes default to "with sharing." Legacy code can silently behave differently — fewer records returned, FLS suddenly enforced. This is the one that bites quietly, so I'd check it before turning on anything new. 

 

2. Pilot ONE agent feature, not all of them. Customer Engagement Agent if 24/7 lead follow-up is the gap; Momentum if CRM hygiene is. One feature, configured properly, measured against a baseline. Breadth is how pilots stall. 

 

3. Hold off on Multi-Agent Orchestration. It's GA and it works, but routing depends on how well each agent is described — which only holds up once your individual agents are clean. Month 2-3, not week 1. 

 

Two questions for the community: 

 

1. What's the first Summer '26 feature you're turning on — and why that one? 

2. Has anyone hit the API v67.0 behaviour change in a real org yet? Curious what broke. 

 

#Agentforce #Salesforceadmins

0/9000

Trailhead Tuesday!

  

Discover how to map agent behavior to conversation design to deliver a seamless user experience with Agentforce. 

 

From this module, you'll learn:

  • The foundations of agent behavior and conversation design. 💬
  • How to design trusted, ethical behavior for agentic AI.
  • Key skills to build reliable interactions that drive customer success.

 

Ready to expand your knowledge? Dive into the badge today to deliver a next-level experience with Agentforce!

0/9000

📣 Session 3 of 8 - Registration Open - [RSVP Click Here] De-Fragging the Franken-Org: Legacy to Agentforce Ready Episode 3: The Data Architecture Lens 📅 Sunday 21 June 2026 ⏰ 11:00 AM Dubai · 8:00 A

 

📣 Session 3 of 8 - Registration Open - [RSVP Click Here

 

De-Fragging the Franken-Org: Legacy to Agentforce Ready 

Episode 3: The Data Architecture Lens 

 

📅 Sunday 21 June 2026 

⏰ 11:00 AM Dubai · 8:00 AM London · 12:30 PM Mumbai 

 

Your Agentforce agent reads your data as truth. 

 

It doesn't have tribal knowledge. 

It doesn't know that "Acme Corp," "ACME Corp," and "acme corp" 

are the same company. 

It doesn't know that the email address on that Contact 

is from 2019 and nobody's used it since. 

 

It reads what's there. And presents it confidently. This Sunday — we fix the data layer. 

 

✅ Data Cloud connection + Identity Resolution 

✅ Golden Record strategy — eliminate duplicates at source 

✅ Einstein Trust Layer — classify and protect PII 

✅ 5 Data Quality Dimensions for AI Readiness 

 

Live. On the Salesforce Trailblazer Community. 

https://trailblazercommunitygroups.com/events/details/salesforce-salesforce-developer-group-dubai-uae-presents-de-fragging-the-franken-org-session-3-your-data-is-a-mess-and-the-ai-knows-it/

 

@Salesforce Developer Group, Dubai, UAE @Agentblazer Community Group @* Salesforce Developers * @WFD Partner Cohort: TDX2630DaysChallenge @Architect Trailblazers #NoClicksJustVibes #DeFraggingTheFrankenOrg

0/9000
0/9000

I have created a new scratch org through the SF CLI and successfully pushed the code/metadata for my Agentforce Agent Topics, Actions, Prompt Templates etc. that I created in a Developer Edition org I created through the Environment Hub and was enabled for Agentforce and Data Cloud pre Agentforce becoming GA.

 

In this new scratch org, I have enabled Einstein under Setup > Einstein Setup, but I don't see any Einstein related Permission Sets etc. I can assign to view the Agents, Agent Actions etc. that I have already pushed using the SF CLI.

 

When creating the scratch org, using the SF CLI, I added "EinsteinGPT" to my project-scratch-def.json, which allowed me to push the Agentforce Agent metadata for Topic, Agent Actions etc.

 

Is it possible to enable Agentforce in a scratch org? If so, what am I missing?

9 Antworten
  1. 22. Nov. 2024, 20:59

    @Mike Sowerbutts have a look at this documentation. It explains everything that needs to be done to enable Agentforce in a scratch org.

0/9000