phi@hanoi:~$
ch.01/07intro
~/growth-marketing · hanoi, vn

i learn by building real things.

An early-career growth marketer who'd rather ship than just study. Across web3, e-commerce and AI, my loop is always the same: try fast, ship real things, learn faster. I care about the business behind the numbers — not just how to grow, but why it lasts.

$ scroll --down
Editorial fashion portrait
building · online
growth · web3 · ecom
0+
ad creatives shipped
~0min
per-creative turnaround (≈5× faster → more hooks tested)
0
community posts / year
02
about

try fast. learn faster.

I'm early in my career, but I don't wait around to learn. My approach is always the same: jump in, build something real, measure what happens, and fix fast.

I learned growth by living it, not studying it. With Monad, I made content, went to meetups, and helped build community culture — and saw firsthand what makes people stick: not features or ads, but feeling part of something.

What I care about most is the business behind the growth — not just pulling users in, but building something that lasts. Every project below is the same loop in a different arena: try fast, learn faster — and prove it with numbers.

03
the work

Experience

Performance Marketing Intern — Saint Gard

apr — jun 2026
ai-assisted creative production · paid media

My most hands-on growth role yet — where try fast → learn faster became a daily loop.

  • Used AI tools to produce 150+ static ad creatives for live campaigns — generating base visuals with AI, then refining styling, layout and final QC by hand.
  • Cut per-creative production from ~2–3 hours to ~30 minutes — at the team's volume, work that would have cost ~375 production hours by hand. The win wasn't speed; it was widening how many hook and format variants the team could test in market (self-measured).
  • My creatives were selected by my manager for the brand's fanpage content and live ad campaigns.
  • Researched 45+ competitor fashion ads, categorized by hook type, visual format and CTA pattern — organized into a reusable swipe file for the team.
  • Tracked daily campaign performance (CTR, CPC, ROAS) and compiled reports to surface top-performing creatives.
creatives selected for live campaigns on saint gard's fanpage
Saint Gard campaign — leather texture Saint Gard campaign — beach editorial Saint Gard campaign — luxury editorial

Content Contributor — Nami

2023
community content · crypto

Where I first learned community-led growth — from the inside, as a user.

  • Published 309 community-engagement posts over the year through a consistent content cadence.
  • Earned "Featured Member" — a points-based recognition driven by content output and engagement (likes, comments, shares).
featured member — sample community posts
Nami community post — DePIN ecosystem analysis Nami community post — BTC cycle analysis Nami community post — Crypto asset flows
04
the shop

concept → live

live product solo build ai-assisted dev

Agent-Care.xyz

may 2026 — present
founder · end-to-end build

Most AI agents are general-purpose. Agent-Care is built on a different premise: agents work best when they go deep on one thing — and the best way to scale complex tasks is to let specialized agents find, hire, and pay each other, without a human in the middle.

Agent-Care is a machine-to-machine marketplace where AI agents autonomously delegate to other agents — no human required in the loop.

True to "system first," I mapped the platform architecture and user journey before writing a line of code — then built and shipped a working product with AI. Currently live on Arc Testnet.

Early signal: 34+ signups in ~1 month through structured Discord outreach, before any real marketing push.

The whole loop in one project: ship the product, watch the first signups, learn what pulls users in.

live product — built end to end
Agent Care — landing page Agent Care — documentation Agent Care — marketplace app Agent Care — agent sandbox
visit live →
e-commerce ops end-to-end shopify

The Vintage Journal

oct — dec 2025
e-commerce ops · shopify

Ran a fashion Shopify store end to end, owning the full e-commerce operations loop — store build, product setup and paid acquisition. I built the store, ran live Meta Ads tests, and learned firsthand what stops the scroll in fashion — which hooks and creative actually drive attention and clicks.

05
on-chain

growth in the wild

Across 35+ web3 token-incentive programs in DeFi, infrastructure, L1/L2, AI and gaming, I've been inside the growth funnels these protocols use to acquire and retain users — quest-based onboarding, referral multipliers, points-to-token conversion, community-led acquisition. A live lab for the exact mechanics modern growth runs on.

06
what's inside

Stack

Run

channels & ops
  • Meta Ads
  • Email marketing
  • Shopify / ecom ops

Make

creative production
  • AI creative tools (image / video generation)
  • Prompt engineering
  • Copywriting
  • CapCut · Canva

Measure

research & data
  • Market & competitor research
  • Google Sheets
  • Notion
07
how i work

the same loop, every time — try fast, learn faster

I map before I move.
When I take on a task, I break it into steps and get clear on what the goal actually is — before I gather tools or resources. Clarity first, motion second.

I work well within a team.
I'm comfortable taking direction, coordinating with others, and owning my part of a shared goal. I like the autonomy to figure out how, as long as I'm clear on the what and the why.

I orchestrate; AI executes.
I split the work into pieces, hand each piece to the right AI tool, then review, refine, and decide what ships. AI does the heavy lifting — the judgment stays mine.

I move fast and learn faster.
I'd rather try, get feedback, and fix quickly than wait for perfect. I take feedback as fuel, not criticism — the faster the loop, the faster I improve.

I'm not afraid of being wrong. I'm afraid of being stuck.
A mistake I learn from is progress. The only failure I avoid is repeating the same one twice.

I keep people in the loop.
I'd rather over-communicate than leave a manager guessing. Clear updates, no surprises.