AI Product Owner

Agent-first highlight
PATI Group is becoming an Agent-first company. Staff get uncapped access to frontier models - not a free-tier hobby setup. Today that includes Claude Code 20x, ChatGPT 20x, SuperGrok Heavy, KIMI Vivace (20x), Google Ultra, and Seedance. We want builders who use the best models in the world as daily infrastructure.
About PATI
PATI is a global DTC e-commerce company selling into the US/EU. We are a lean, high-performance team based in Ho Chi Minh City, restructuring and strengthening the org to become an AI Native company: AI is operating infrastructure, not a side tool. We use AI to cut repetitive work; people keep understanding the problem, checking results, and owning decisions.
Three things that shape how we work:
Lifelong learning. If you believe in lifelong learning, PATI fits. Signals we look for: you have paid for professional courses yourself, or you learn regularly through YouTube, X and social, or you read professional books consistently. Learning fast is a real advantage here.
AI First. AI helps you move faster, but the first core capability at PATI is still thinking. AI amplifies your capability; it does not replace your responsibility to think. Fit: you like finding ways to use AI to raise work quality, and when you see the standard is not met, you redo it. Not a fit: depending on AI, pasting answers to colleagues when you do not understand them yourself, or sketching more than shipping.
High performance. If unfinished, below-standard work bothers you, you keep time, you optimize how you work, and you measure your own output, PATI fits. Decisions here are based on results and data, not feelings. We evaluate by measurable results, not hours at a desk - that is why compensation is reviewed 4 times a year.
Scope of work
PATI Group is not looking for a passive coder who waits for tickets. We are looking for a Product Manager - the product owner for several product positions under PATI OS: AdLauncher, StoreLauncher, Creative Portal, AdVault, AI Portal, and Profit Center.
You turn department pain points into shipped internal products, hand them to non-technical users, and keep upgrading them against P&L. You will not ship every surface on day one - sequence with 80/20.
40% - Multi-app product ownership (PATI OS)
Outcome: each assigned product has a clear problem, roadmap, Definition of Done, and live users who can run it without you.
Concrete work:
Own briefs and roadmaps across AdLauncher, StoreLauncher, Creative Portal, AdVault, AI Portal, and Profit Center as capacity earns them.
Sit inside Marketing / Creative / Store / Ops / Finance workflows. Dig the real bottleneck - do not wait for a perfect ticket.
Sequence build vs buy vs kill. Cut features that do not move revenue, creative throughput, or cost.
Write requirements engineers and AI builders can execute: user, job-to-be-done, acceptance tests, failure modes.
Ship in thin slices. Demo to real users. Measure adoption. Run v2 from feedback - "1% better every day."
25% - AI workflow & system design
Outcome: flows are repeatable and adoptable by non-technical staff (handover is part of Done).
Concrete work:
Design prompts, system instructions, agent roles, MCP/CLI hooks, and context/memory so models execute reliably across PATI OS surfaces.
Prefer typed judgments and agent loops where they beat chat-and-paste.
Instrument products: usage, failure rates, time saved, quality gates before output hits ads, creators, or digests.
Partner with CTO / builders on architecture; you own the product outcome, not only the backlog list.
20% - Discovery, prioritization & P&L alignment
Outcome: every build ties to revenue, creative capacity, or measurable cost - not lab demos.
Concrete work:
Prioritize with Managers and the Board using 80/20 across the PATI OS portfolio.
Translate technical options into business language for Leaders; argue toward clear requirements.
Track product metrics that matter: adoption, cycle time, rework, paid-tool replacement, quality escapes.
Kill or park work that does not clear the bar. Say no with evidence.
15% - Rollout, training & continuous upgrade
Outcome: products stay alive in production after launch.
Concrete work:
Train internal users; write short playbooks; own onboarding for each release.
Collect feedback, triage bugs vs enhancements, publish a visible changelog.
Keep AdLauncher, StoreLauncher, Creative Portal, AdVault, AI Portal, and Profit Center coherent under one product standard.