Oracle AI Agent Studio Workflow Agents Use Case#2: Bulk Replace Project Manager across Projects

This is Episode 7 of the Workflow Agents: A Complete Guide series, where I build a custom AI Project Resource Replacement Agent from scratch inside Oracle Fusion AI Agent Studio. 🔁 Replace a project manager across ALL assigned projects in Oracle Fusion PPM — using just one natural language command. No manual updates. Just AI. 💡 This use case came directly from the Oracle Community — a colleague asked whether Oracle would ship a standard agent for this. Instead of waiting, I built one. ──────────────────────────────────────── 🧠 KEY CONCEPTS COVERED ──────────────────────────────────────── Oracle AI Agent Studio — Workflow Agent design BI Publisher SOAP integration as External REST Tool Base64 decoding via LLM prompt engineering Parallel Processing For Loop node for multi-project processing HCM Worker REST API (expand=emails, field filtering) Project Team Members — PATCH (end-date) + POST (create) LLM-driven payload generation with date validation logic Context variable passing across nested loop nodes Production deployment: agent publishing + role-based access ──────────────────────────────────────── ⏱️ TIMESTAMPS ──────────────────────────────────────── 00:00 – Introduction & Episode Overview 00:46 – Community Use Case Story (Why I Built This) 01:25 – Process Flow Walkthrough (4-Stage Architecture) 05:33 – Demo Begins: Inside Oracle AI Agent Studio 06:11 – Why Workflow Agent (Not Supervisor Agent) for This Use Case 07:50 – Creating the Workflow Agent (Name, Family, Type, Variables) 09:57 – Node 1: Extract Resource Entities (LLM — NLP Parsing) 13:00 – Node 2: Initialize Variables (Set Fields Node) 14:17 – Node 3 & 4: Fetch Outgoing & Incoming Worker (HCM Business Objects) 20:55 – Node 5: Generate Project Report (External REST Tool — BIP SOAP Call) 28:43 – Node 6: Parse Report to Array (LLM — Base64 Decode & JSON Transform) 33:39 – Node 7: Process Projects Loop (For Loop — Parallel Execution) 35:57 – Business Objects: Project Search + Project Member Search 40:44 – Business Objects: Update & Create Project Team Members (PATCH + POST) 48:58 – Node 10: Calculate Swap Payload (LLM — Date Logic + Dual JSON Output) 54:00 – Date Validation Logic Explained (System Date vs. Project End Date) 58:43 – Testing Payload Generation: Incoming + Outgoing Payload Verification 01:07:28 – Node 13: Generate Project Summary (Per-Loop LLM Summary) 01:09:11 – Node 14: Compose Final Notification (Aggregated User Message) 01:10:27 – End-to-End Test Run: Lawrence McCord → Kathy Anderson (1 Project) 01:13:43 – Verifying Results in Oracle Fusion PPM 01:15:43 – End-to-End Test Run: Frederick → Carl Anderson (2 Projects — Loop) 01:16:47 – Verifying Multi-Project Swap in Oracle Fusion 01:21:29 – Production Readiness: Security, Roles & Publishing the Agent 01:22:04 – End-User Experience: AI Chat Window (Published Agent) 01:22:57 – Closing Thoughts & Next Steps ──────────────────────────────────────── 📌 RESOURCES & LINKS ──────────────────────────────────────── 📝 Community Link for use case: https://community.oracle.com/customer... 📂 Git Hub repository for full details: https://github.com/arpitpaliwl/Oracle... ──────────────────────────────────────── 📝 Submit Your Oracle AI Agent Use Case: Have a use case you want covered in this series? Fill this quick form 👉 https://forms.gle/kyShoQjV875w5PFi7 🔔 Subscribe for the complete Oracle Fusion AI Agent Studio series — covering workflow agents, supervisor agents, agentic apps, evaluation, and real-world use cases. 👤 About Me: I'm Arpit Paliwal — Oracle ACE and Certified Oracle ERP/AI Consultant. I create practical, hands-on content on Oracle Fusion, AI Agent Studio, and enterprise automation. Follow me for more: 🔗 LinkedIn:   / arpitpaliwal11   🔗 Substack: https://arpitpaliwaloracleaierp.subst... #oracleaiagentstudio #oraclefusion #Oracleppm #workflowagents #oracleerp #aiagentstudio #enterpriseai #oracleace #LearnWithArpit #oraclecloud #projectmanagement #erp #digitaltransformation #generativeai #agenticai