Skip to content

Quick_Start_2Pager.md

Latest commit

 

History

History
35 lines (26 loc) · 1.39 KB

Quick_Start_2Pager.md

File metadata and controls

35 lines (26 loc) · 1.39 KB

Agentforce Data‑Aware Solutions — Quick Start (Exec & PM 2‑Pager)

Why: Faster ops, no manual field mapping, governed (FLS, deny‑list), auditable.
Pilot metric ideas: +30% triage speed, +15% timely follow‑ups, <1% permission violations.

Scope (4–6 weeks)

Two complementary AI agent solutions:

Agentforce Data-Aware Agent

  • Use case: Lead qualification & follow‑up automation
  • Surfaces: Lightning / Slack (optional)
  • Data: Lead (+ related minimal fields), PII deny‑listed

Personal Shopping Assistant

  • Use case: AI-powered e-commerce customer service
  • Surfaces: Einstein Bots, customer portals
  • Data: Product catalog, inventory, customer preferences

RACI (pilot)

Sponsor (A), PM (A), Admin (R), Dev (R), Security (C), Sales Ops (C), E-commerce (C).

Environments

Dev Hub + Scratch (CI on PRs) → UAT Sandbox → Production

Checklist

Week 0: GitHub secrets; deny‑list review.
Week 1: Scratch, deploy, bootstrap, smoke test.
Week 2–3: Tune prompts/actions; iterate.
Week 4: UAT metrics; approve; Prod.

How it works (plain English)

Planner finds objects/fields/paths from a cached schema graph, then calls safe queries/Flows/Apex under user permissions. Only a small schema slice is sent with prompts; sensitive fields are blocked.

Governance & risk

FLS enforcement; RestrictedField__mdt; audit stubs; rollback via PRs/tags.