What I do

👋 Hi, I'm Mia Zheng —

I ship GTM. I build AI agents. I write about both.

Seven years in SaaS, and I'm still chasing the same loop: demand gen, activation, retention — that's where revenue actually happens. This is where I share what I'm learning about running it with a small team plus a team of AI agents.

Scroll down for how to reach me

B2B SaaS GTM Lead who ships strategy and code.

Shopify Ecosystem Expert 🤖Multi-Agent AI Orchestration (n8n) 📈APAC, UK & US Market Expansion

Receipts · Numbers · Evidence

Numbers don't lie.

Seven years of GTM and growth work, distilled into three results. Each one earned, measured, and shipped — not theoretical.

Flagship Result
$0M

First dollar to $1M ARR in 24 months — no dedicated sales team, owning every GTM call end to end.

User Growth
+0%

In 3 months. Lean budget. More users, lower CAC. No waste.

AI Efficiency
−0%

Manual B2B prospecting time, cut by an AI-powered lead tool I built.

AI-Native GTM · How I Actually Work

I read the market with AI.
Then I run on it.

Sensing and shipping are the same job — most teams only staff one half of it. I use Claude in Chrome to read where the market is actually moving, then hand that read to a five-agent system I built in Codex: competitive analysis, positioning, campaign copy, visuals, and the report at the end.

It’s not a stack of tools. It’s an operating system I run.

Agent Network · n8n Live Standby
GTM Manager
Brain · Dispatcher
Scanning competitors
Research
Intelligence
Drafting Twitter thread
Content
Copywriter
Generating visuals
Image Gen
Creative
Compiling report
Visualize
Analyst

5 GTM Mistakes Everyone’s Making

Claude in Chrome, pointed at the LinkedIn GTM conversation. A scroll of scattered takes in, five patterns out. EN · 1:49

Also on YouTube ↗

Five agents. One brief in, a campaign out.

Research, copy, visuals, and reporting — handed between agents on a loop that closes without me in it.

I break these workflows down as I build them. More on the two places I post.

Fresh off the blog

What I'm thinking about this week.

Notes on GTM, growth, and building with AI agents — published every 2–3 days.

Shipped · Public

Strategy decks are opinions.
This one ships as code.

The system above runs my own GTM. This one I gave away. I kept rebuilding the same deck by hand, so I built the skill that stopped me having to — then put it on GitHub for anyone stuck in the same loop.

github.com/miazheng-max

show-anywhere-slide

A slide-generation skill for Claude Code and Codex CLI. You pick the look by seeing it — a browsable gallery of real swatches in real display fonts — instead of describing your taste in two adjectives. Portrait-first, because most decks get read on a phone now.

46design systems
0dependencies
1self-contained file
9:16default canvas
View the repo
Style Gallery — pick by sight, not by adjective
The style gallery: 46 design systems shown as real cards in their real display fonts
Real output — 9 of the 46 bundled styles.

Let’s talk

If you’re building or writing about similar things, I’d love to hear from you.

Whether you’re figuring out GTM in a small team, playing with AI agents in your workflow, or just want to say hi about something I wrote — happy to compare notes.

miamoneyzhaocai@gmail.com
Mia Zheng Hangzhou · Writing weekly