The newsletter for freedom-seekers

This week: How Tony Dinh went from side projects nobody used to $130–160k/month with TypingMind — the weekend project that accidentally became his life's work. Plus the two failed AI products that nearly killed his confidence, and the one move that changed everything. 

EmojiAI made $0.
AskCommand made $100.
TypingMind made $1,500,000.

$130–160k monthly revenue 4 years to get here Da Nang → World 1 person no employees

In August 2021, Tony Dinh walked out of his software engineering job in Singapore for the last time. His salary had been good — around $105k a year. He had spent seven years building other people's products. He had also spent two years building his own on the side, quietly, without telling his colleagues.

He had enough savings to survive four years in Vietnam without revenue. His goal was modest: $1,000 MRR in the first year. Enough to live comfortably. Enough to prove the idea was real. He went home to Vietnam, opened his laptop, and got to work.

The next two years were not a straight line. His first big product, Black Magic — a Twitter analytics tool — grew steadily to $13k MRR. Then Twitter changed its API pricing to $42k/month overnight and wiped out most of the market. Tony sold the product for $128k rather than watch it die. He also built two AI products — EmojiAI and AskCommand — that made a combined total of around $100 despite one launch tweet getting 5,000 likes.

❝

"I went outside for a walk, took a fresh breath of air, and felt the freedom I never experienced before."

Then, on March 1st, 2023, OpenAI released the ChatGPT API. Tony registered typingmind.com that same day — the domain name suggested by ChatGPT itself. He spent four days building an MVP. He shipped it. Within seven days he'd made $22,000. Within a year, TypingMind crossed $1M in annual revenue. By late 2025, it was making $130–160k a month — over 50% of it recurring B2B revenue from the Team version.

Tony now travels the world with his wife, surfing in Bali, eating in San Sebastian, coding from cafés in Da Nang. He works about four hours a day. He's reached what he calls financial independence — not just income, but the option to stop entirely if he wanted to. He doesn't want to. That, he says, is how he finally figured out it's actually what he loves.

The Teardown
Business model
SaaS — one-time + subscription
Personal license + B2B Team plan. Recurring now exceeds 50% of monthly revenue
Revenue
$130–160k/month
$1M+ annual. Solo operation, no employees, ~90% margin
Top traffic source
Word of mouth
Paid ads produced 2 conversions on $600 spend. Product quality drives organic growth
Monthly expenses
Minimal
Lifestyle unchanged since 2021. Da Nang cost of living + modest travel. No office, no team
01
He launched in four days, not four months. When OpenAI released the ChatGPT API on March 1st, Tony registered a domain that same day and shipped an MVP six days later. He didn't wait for a perfect product — he shipped something that solved his own annoyance with ChatGPT's slow interface and missing search. Speed of shipping was the entire moat.
02
The B2B pivot is what made it a real business. The personal license was a great launch, but the Team version — a white-label enterprise deployment — is what pushed revenue past $1M/year. The shift from one-time purchases to recurring B2B contracts gave him stability he'd never had with Black Magic or his earlier products.
03
Two years of building in public before TypingMind existed. Tony spent 2020 and 2021 tweeting, writing newsletters, and sharing his journey publicly — growing to 50k+ Twitter followers before TypingMind launched. That audience didn't guarantee sales (see: EmojiAI), but it meant a built-in distribution channel when the right product arrived.
04
The mistake he almost made: staying dependent on one platform. Black Magic's collapse when Twitter changed its API was a brutal lesson. TypingMind was deliberately built to be model-agnostic from day one — it works with OpenAI, Anthropic, Google, and others. No single API provider can kill it the way Twitter killed Black Magic.
The Steal
Three things from Tony's playbook you can copy today:
01
The launch post formula. Tony's Product Hunt and Twitter launches follow the same 5-line structure: one sentence on the problem, one on the solution, one on what makes it different, a demo link, and a direct ask ("what do you think?"). No preamble, no features list. He registered typingmind.com and tweeted an MVP in six days — the tweet was the product.
02
The build-in-public update template. Every feature Tony ships becomes a public post structured the same way: what he built, why he built it (his own frustration), and one specific number (days to build, revenue, users). The frame is always "here's what I learned" not "look how great this is." That framing is why people share it.
03
The platform-risk checklist. Before building on any API or platform, Tony now asks: can this provider change pricing and kill my business overnight? If yes, can I swap them out in under a week? TypingMind works with OpenAI, Anthropic, and Google — no single provider can do to it what Twitter did to Black Magic. Run this check on everything you're building on right now.
THIS WEEK’S MOVE

Find your API dependency before it finds you

If you're building anything on top of a platform you don't control — Twitter, OpenAI, Shopify, App Store, Google — you have a Black Magic problem waiting to happen. Tony learned this the hard way. You don't have to.

This week's move takes 45 minutes and doesn't require you to build anything. It's a risk audit. The goal is to find your single biggest dependency and make one concrete plan to reduce it.

  1. List every external platform your revenue depends on. Not just APIs — include App Store distribution, Google search traffic, a single social platform, Stripe, any marketplace. Write them down. If any one of them changed their pricing or terms tomorrow, which one would hurt you most?

  2. For your top dependency, find one alternative that exists today. You don't have to switch — just verify an escape route exists. If you're on OpenAI, does your product work with Anthropic or Google? If you rely on App Store, can you sell direct? Knowing the exit makes you less trapped.

  3. Add one email capture anywhere in your current product or content. Tony's newsletter and Twitter following meant he owned a direct line to his audience no platform could take away. An email list is the only audience you actually own. If you don't have one, start one this week — even just a simple landing page.

Nomad Intel
Visa
Vietnam's Golden Visa is still a proposal. No dedicated nomad visa as of March 2026. Most founders use the $25 90-day e-visa and do quarterly border runs to Bangkok or KL ($200–400/trip). For stays over a year, a business (DN) visa via a local sponsor is the cleanest route. Thailand's DTV is ahead on legal clarity — worth comparing if you're planning a longer base.
City
Da Nang just had its second Nomad Fest (March 20–29). Coworking day passes from $3, cost of living $800–1,200/mo. Growing fast — which means the An Thuong strip is getting noisy and expensive. Son Tra neighbourhood is the quieter call. Best for: Southeast Asia affordability with an actual indie community attached.
Gap
The AI chat interface market is full. Client management isn't. Lightweight tools for solo founders — contracts, proposals, async client comms, project delivery — remain underserved. Existing options are either too enterprise (Dubsado, HoneyBook) or too DIY (Notion). Freelance management is a $9.2B market by 2030. One person who lives this problem could own it.
Want to follow Tony's journey yourself? He publishes monthly updates on revenue, travel, and building in public.
Read Tony's newsletter →
BEFORE YOU GO:

TypingMind — typingmind.com | The product we just tore down is genuinely useful. If you're paying for ChatGPT Plus but frustrated by its interface limitations, TypingMind gives you a better UI across multiple models for a one-time fee. Worth trying before you form an opinion on Tony's story.

That’s it for this week. Build something that travels with you. 🌍

P.S. Forward this to a fellow nomad (or wannabe nomad) who dreams of running their own thing.

Keep reading