Senior Product Manager – AI Platforms \& Agentic Workflows
About the Role
You will lead the modernization of enterprise workflows by deploying agentic processes on top of an enterprise LLM platform and a dedicated agent-orchestration layer. Your work will drive measurable improvements in cycle time, right-first-time quality, and rework—while building a durable enterprise knowledge base (KB).
Main Duties \& Responsibilities
1. Product Discovery \& Workflow Design
Identify, map, and prioritize high-value workflows for automation.
Develop PRDs with clear KPI targets, success criteria, and evaluation methods.
2. Agentic Orchestration
Design planner/executor/verifier workflows and human-in-the-loop (HITL) gating.
Manage agent versioning, evaluation frameworks, guardrails, telemetry, and rollback plans.
3. LLM Platform Management
Oversee LLM platform governance including access, policies, security, and evaluation.
Maintain clear separation between the LLM foundation layer and agent-orchestration layer.
4. Integrations \& Data Enablement
Coordinate with ERP/CRM/ITSM/SCM/QMS and other systems via APIs and eventing.
Implement RAG/search components with proper RBAC, observability, and audit trails.
5. Enterprise Knowledge Base (KB)
Establish taxonomy, ontology, editorial rules, ingestion workflows, and write-back processes.
Ensure agents both use and enrich the KB to improve long-term accuracy.
6. Measurement, ROI \& Cost Control
Baseline operational metrics and build dashboards for performance monitoring.
Manage token budgets, latency constraints, and AI unit economics.
7. Change Management \& Adoption
Update SOPs/runbooks; train and enable process owners.
Drive adoption through demos, job aids, and feedback loops.
8. Stakeholder \& Vendor Management
Partner with Legal, Security, Finance, and cross-functional leaders.
Work with LLM/agent vendors while maintaining a vendor-agnostic approach.
Key Outcomes (90–180 Days)
30–45 days:
Map 6–10 workflows; baseline KPIs; build ROI model; publish P0 roadmap.
60–90 days:
Deliver 2–3 production-grade agents with telemetry and HITL guardrails.
By 180 days:
Achieve KPI improvements in ≥3 workflows; launch KB 1.0 (taxonomy \& governance).
Minimum Requirements
4–8+ years
in Product Management, Platform PM, TPM, Process Owner, or related roles.
Demonstrated experience in workflow/process redesign and KPI measurement.
Strong understanding of system integrations (APIs/eventing, SSO/identity, logging/metrics).
Working knowledge of AI/automation technologies (LLM/ML/analytics/RPA).
Excellent communication skills in
Vietnamese and English
.
Proven ability to influence stakeholders and lead change in complex environments.
Preferred Qualifications
Experience with agentic patterns: planner/executor/verifier, HITL, evaluation frameworks, rollback.
Background in operations transformation (manufacturing, logistics, BFSI, customer operations).
Experience managing AI cost models (token budgets, latency SLAs, unit economics).
How We Work
Evidence-based:
Every release ships with metrics, evaluations, and rollback plans.
Human-in-the-loop first:
Agents augment humans; humans own outcomes.
KB-centric:
“If it’s not in the KB, it’s not done.”