Trailblazer Community
Feed

Getting Started

  • What is Trailhead
  • myTrailhead
  • Trailblazer Community
  • Career Growth

Learn

  • Trails
  • Trailmixes
  • Modules
  • Projects

Credentials

  • Superbadges
  • Certifications
  • Maintain Certifications

Salesforce

  • Salesforce.com
  • AppExchange
  • Developers
  • Help Portal
© 2026 Salesforce, Inc. All rights reserved. Privacy Terms of Use Trust Accessibility Cookie Preferences

Feed

Browse what's being discussed across the community. Join the conversation in Slack.

  • AS
    Akash Sharma
    Salesforce Administrator · 2 hrs ago
    Asked in Agentforce Builders

    Best pattern for handing off from Agentforce to a human agent mid-conversation?

    My sales agent is taking the first turn of every inbound chat, but I want a clean way for it to flag complex deals and pull in a human AE without losing context. Has anyone wired this up with Omni-Channel routing yet?

    Accepted answer
    MI
    Mia Iyer
    Senior Architect · Salesforce MVP · 1 hr ago

    Capacity model + a custom skill for 'AE-routable-deal' is what's worked for us. We don't keep the agent in the conversation mid-stream — it posts a structured handoff summary as the first message in the new conversation, and the AE picks it up cold with the full context attached. Cleaner than 'silent watching'.

    14 replies
  • SV
    Sofia Vasquez
    Senior Content Author · 4 hrs ago
    Asked in #Agentforce

    How are you grounding Agentforce responses without leaking PII into the prompt?

    We're piloting Agentforce for inbound case triage. Need to ground the model on knowledge articles + recent case history, but our compliance team flagged that the unfiltered case history can include unredacted PII.

    Accepted answer
    MI
    Mia Iyer
    Senior Architect · Salesforce MVP · 2 hrs ago

    Two-stage retrieval has been the cleanest pattern: pull the case history through a redaction layer first (the Einstein Trust Layer handles a lot of this by default but you can extend it), then ground the agent prompt on the redacted versions. Your compliance team should be able to approve the redaction policy once, not every prompt.

    19 replies
  • MI
    Mia Iyer
    Senior Architect · Salesforce MVP · 5 hrs ago
    Asked in #Data Cloud

    Identity resolution rulesets — when to use deterministic vs. probabilistic?

    We have ~40M unified profiles and the dedupe rate is lower than I expected. Walking through whether probabilistic blending makes sense for the consumer side of the org.

    Accepted answer
    JY
    Jasmine Young
    Senior Salesforce Administrator · 3 hrs ago

    We ran the same comparison last quarter. Probabilistic was the right call for us but we capped the confidence threshold high — 0.95+ only. Caught the household collapse problem you're worried about by adding a 'shared-address-but-different-DOB' exclusion to the ruleset.

    27 replies
  • JB
    Jordan Brooks
    Account Executive · May 16, 2026 11:42 AM
    Asked in Service Cloud Voice

    Service Cloud Voice transcripts as Agentforce prompts — anyone tried it?

    We're sitting on years of Voice transcripts. Wondering if there's a clean way to use them as few-shot examples for an Agentforce intake agent without rolling our own pipeline.

    6 replies
  • JY
    Jasmine Young
    Senior Salesforce Administrator · May 16, 2026 9:12 AM
    Asked in #Flow Builder

    Process Builder → Flow migration: cross-object field updates are biting us

    The migration tool generated a flow that loops over child records but the order of execution shifted vs. the old Process. Sharing what we did to recover.

    8 replies
  • MB
    Maya Bennett
    Salesforce Admin · Group Leader · May 15, 2026 8:30 PM
    Asked in London Trailblazers

    Recap: London Trailblazers meetup — Agentforce night, 60+ admins

    Big thanks to everyone who came out. Slides and the demo org link are in the channel. Next month: Data Cloud + Tableau Pulse.

    5 replies
  • EO
    Emeka Okonkwo
    Salesforce Administrator · May 14, 2026 2:45 PM
    Asked in #Apex

    Apex bulkification — is the "one trigger per object" rule still load-bearing in 2026?

    Newer trigger handler frameworks split things up by feature. Curious if anyone has pushed back on the classic single-trigger guidance.

    41 replies
  • AG
    Alex Gomez
    Account Executive · May 14, 2026 10:15 AM
    Asked in #Tableau

    Tableau Pulse rollout — how are you handling permission sets for analyst tiers?

    Spinning up Pulse for ~600 users across three analyst tiers. Defaults don't quite match how we'd carve up data access. Anyone with a clean tier-based pattern?

    7 replies
  • EO
    Emeka Okonkwo
    Salesforce Administrator · May 13, 2026 5:30 PM
    Asked in #Developer

    LWC shadow DOM gotcha when nesting an LWR-rendered component inside a LWC

    Hit a styling issue where the parent LWC styles bleed into the LWR component via slotted content. Curious if anyone has a reliable pattern for this.

    11 replies
  • AS
    Akash Sharma
    Salesforce Administrator · May 13, 2026 2:20 PM
    Asked in Agentforce Builders

    First Agentforce deployment — pre-flight checklist you wish you had?

    Going live with our first sales agent in two weeks. Beyond the docs, what's the unwritten checklist you wish someone had given you before deploy day?

    23 replies
  • Previous
  • 1
  • 2
  • 3
  • Next
Salesforce Featured

Conversation lives in #community. Open Slack to post, reply, and react.

Trending topics

  • Agentforce1,842
  • Apex980
  • Data Cloud612
Browse all topics →

Active groups

  • Agentforce Builders
    12,480 members · #agentforce-builders
  • Service Cloud Voice
    5,210 members · #service-cloud-voice
  • London Trailblazers
    2,310 members · #london-trailblazers
Browse all groups →

Top contributors this week

  • MI
    Mia Iyer
    All-Star Ranger
  • JY
    Jasmine Young
    All-Star Ranger
  • MB
    Maya Bennett
    Mountaineer
  • EO
    Emeka Okonkwo
    Adventurer
  • SV
    Sofia Vasquez
    Ranger