Running My Apparel Brand on AI Ops by Hai BuiRunning My Apparel Brand on AI Ops by Hai Bui

Running My Apparel Brand on AI Ops

Hai Bui

Hai Bui

About the work

I founded Ky Lan Athletics, a DTC activewear brand, and run it with a deliberately small team — a virtual assistant and a marketing manager. The way we punch above that headcount: an AI operations layer with three scheduled check-ins a day, a weekly retro, and a shared memory that never loses the thread.

What runs every day

Morning Feed — before I've opened my laptop: today's schedule, yesterday's unfinished items carried forward, and every message across 9+ channels (email, Slack, WhatsApp, Instagram, LinkedIn, Discord, and more) triaged into one scannable report — each one summarized, with a reply already drafted in my voice, matched to that specific person's register — across a team that spans Seattle, Vietnam, and Hungary, with Vietnamese threads drafted natively for my staff in Vietnam. Nothing sends without my approval.
Noon Triage and Evening Wrap-up — two more incremental sweeps, so nothing sits unanswered for more than a few hours. The evening pass also recaps what went out and what's still open, and pre-fills my daily review from the tasks I actually completed.
Daily planning that respects reality — the AI asks how much time and energy I have today, then sizes the day's plan to fit it: priorities ordered P0 to P4, must-dos capped at ~75% of my stated hours, everything synced to my task tracker automatically.
Weekly Retro — every Friday the AI reviews the full week (tasks, journals, decisions), proposes next week's priorities, and maintains its own context documentation, so the system improves instead of drifting.

The part that compounds: a system that learns

Every workflow reads from one living context page — standing rules, active projects, who's who, how I write to each person. When I correct a draft or change a rule, it's written down once and every future session honors it. Months-long negotiations with my manufacturer in China, invoice disputes, event logistics: the AI picks each thread up with full history, every time.
I also built custom skills on top — like an analysis workflow that turns any article, meeting recording, or video I share into a five-section brief tied to my actual business: what it says, what it means for my brand, where it's wrong, and what to do about it this week. Learning material stopped piling up unread.

Why it matters

The real value isn't replacing people — it's operating a level up without a level up in payroll. The repetitive layer is handled before any of us touches it, so a three-person team spends its time on product, marketing, and growth instead of inbox triage and status pulls. Message backlogs became a ten-minute review, three times a day.
This is the exact architecture I install for other DTC brand owners: same system, adapted to their tools, their team, and their voice.
The full system stacks three layers:
The assistant you didn't hire — this case study: the daily grind handled before anyone logs in
The analyst you didn't hireMy Store's Insight System: the compass — what to fix, what to highlight, which ads to run
The developer you didn't hireAI-Ready Shopify Development with Claude Code: fixes shipped in hours, not invoices
Like this project

Posted Sep 3, 2026

Three scheduled AI check-ins run my day at Ky Lan Athletics: every channel triaged with replies drafted in my voice, my task list sized to the energy I actually have, and a weekly retro that keeps the whole system honest. Operating a level up, without a level up in payroll.