First month as a solo builder after leaving big tech
- YT :: https://www.youtube.com/watch?v=GQ0U_gTYMpA
- Original title :: First Month After Quitting My Big Tech Job as L8 Principal
Kun Chen, an L8 principal engineer at Meta (previously Microsoft and Atlassian) with almost two decades in big tech, quit his job to become a solo builder. He explains why, how he weighed the financial trade-offs despite the math not favoring the move, and what the first month of total autonomy actually looked like.
Why he quit
Three converging reasons. First, as he grew more senior, his time became better spent directing others than building himself, but he missed the hands-on process of building, not just its output. Second, AI made it possible for a single person to build far more than before, which he compares to another industrial revolution and made this feel like the right moment to explore ideas solo. Third, personal: his son turned eight, is more independent, and the family had enough savings to sustain a stretch of zero income.
The math didn't favor it, but he did it anyway
He explicitly modeled expected financial outcomes of solo building over one to three years, using four typical outcomes: getting nothing, a soft landing/acqui-hire, a successful exit, or a unicorn. He assigned rough probabilities (most startups get nothing, a moderate share land softly, few get real exits, very few become unicorns) across conservative, mid, and optimistic scenarios. Conclusion: even the optimistic case returned less than three more years at big tech. He decided the math shouldn't be the deciding factor for a life decision like this — the question he asked himself was whether he'd be satisfied looking back having never given himself a shot at building what he wanted, and the answer was no.
First days: learning to use freedom
The first Monday, he had no messages, no meetings, no backlog — a blank canvas, which he found disorienting and had to learn how to use. He started with an idea he'd had strong conviction about for years: an AI tutor for his son (who is learning math, chess, and coding), believing AI could give a much more personalized learning experience than passive video.
Moving fast and the unexpected side quests
With zero organizational overhead, he moved extremely fast, which surfaced new bottlenecks in his own development workflow that no mainstream tools solved. He built tools to fix his own friction and open-sourced them, accidentally becoming an open-source maintainer as a side effect — not his main goal, but valuable both as new experience and as a way to move faster himself. He shipped 30 to over 100 well-tested, documented PRs a day, releasing multiple projects (Good Night Have Fun, an agent orchestrator; Axi; Autopresso; Lavish Editor; Simple Words; and more) alongside the main AI tutor project. He describes the bottleneck shifting to purely the speed and quality of his own thinking — ideas going from conception to live and in users' hands within about 15 minutes, with bug fixes ready before users finished describing the problem.
None of these side projects make money, which he says is fine — right now the focus is building, shipping, and learning what being a solo builder means, not monetization.
Traction and what drove it
He went from near-zero to over 8,000 X followers, Good Night Have Fun organically crossed 1,000 GitHub stars, and he did a three-hour podcast with "Steve" that passed 20,000 views. He draws three lessons from this early momentum:
- Authenticity as a human matters — his AMA thread about quitting unexpectedly went viral (200K views, 300+ questions answered), which he attributes partly to being a real human voice amid generic AI-generated content.
- Craftsmanship matters — he deliberately gave Good Night Have Fun a "personality" (name, starry-night visual theme reflecting its use case of running agents overnight) rather than building a plain CLI tool, and users repeatedly cited the polished experience as what first grabbed their attention.
- Connections matter — his relationship with Steve, formed years earlier from unrelated career-advice YouTube videos, led directly to the podcast collaboration and further introductions to like-minded builders, reinforcing that networking pays off unpredictably even for an introvert.
The double edge of autonomy
Full solo autonomy — beyond what even a co-founder or small founding team would give him — lets him pivot freely (change direction daily, kill an idea without losing credibility, spend a day on a YouTube video instead of "the business"). But he flags this as dangerous: zero accountability makes it easy to scatter across many unrelated efforts without producing a strong outcome.
His fix, drawn directly from big-tech experience: define a clear personal mission, the same way focused companies avoid unrelated ventures (Microsoft doesn't do real estate, Apple doesn't sell coffee, Stripe doesn't build cars). After reflection, he settled on: to help empower every individual to discover, pursue, and achieve their full potential. He now filters every potential project through whether it ladders up to that mission, rejecting ideas that don't, regardless of how appealing they seem.
Advice for others considering the leap
He explicitly says not to copy his path. Instead: be honest about the trade-offs, know what you're optimizing for, and make sure you're running toward something rather than away from something. He plans a follow-up video on his agentic engineering workflows that let him move so fast.