10 years ago I made a decision that changed my life.
When I quit my finance job to teach English in South Korea, I asked myself 4 questions:
1. What's the worst that can happen?
I'd spend a year abroad, learn a new language, and come back with a completely different perspective. Even the "worst case" would push me out of my comfort zone and force me to grow.
2. What's the best case scenario?
I'd break out of the conveyor belt my peers were on (banks, hedge funds, same path). I'd actually expand my horizons instead of settling into an office at 21.
3. Is this reversible?
Finance jobs weren't going anywhere. I could always go back to the traditional path if it didn’t work out. But the window to take risks before life gets complicated? That was closing.
4. Will I regret this?
Definitely not! I'd regret playing it safe rather than taking the risk.
That decision led to everything: a university scholarship, an MBA, meeting my wife, corporate training for Hyundai, building an AI education business in Thailand, working in customer success at @TheRundownAI and now at @Chatbase as the first sales hire.
Most “risky” decisions aren’t that risky when you actually break them down.
One thing I admire about the suburbs is that people have unprecedented pride in their gardens.
The way they spend their time, evenings, and weekends growing plants and doing landscaping doesn't make sense at all to a working professional, because you could be spending that time vibe-coding, selling, or building.
I think they get the same fulfillment from showing off a beautiful garden that I get from closing enterprise deals.
Definitely something to look forward to later on in life.
everyone is arguing about which model is best
wrong argument
the question isn't "GPT vs Claude vs Llama"
the question is "can you switch between them without rebuilding your product"
if the answer is no you've already lost
model neutral is the new cloud neutral
remember when everyone said "don't build on one cloud provider"
same thing is happening with LLMs right now and most people aren't paying attention
model neutral is the way forward
the best part of sales at @chatbase isn't the close
it's when a customer emails you back saying "we went live in 48 hours and our team finally has breathing room"
that's the value right there
not the contract. not the credits. not the SLA
someone's workday got better because of something you built
that never gets old
it's like @Tesla FSD saving you countless hours on the road and providing a safe ride
Surgeons don't wing it. Your AI agent shouldn't either.
Introducing Procedures. You define the trigger, the steps, the branches, and when to hand off to a human. Your agent does the rest, exactly how you designed it.
Live now on Chatbase.
Contains graphic scenes.
11:40pm. An empty office and a whiteboard I’d rewritten four times.
We were planning the quarter forward. Week one, week two, week three. Every version collapsed by row five.
A guy named Dan drew a line down the middle and wrote the number we needed at the far right of the board. Then he asked one question: “What has to be true the day before that?”
We answered it. Then we asked it again. And again.
Twelve steps back, we landed on Monday. Monday had one task on it. One call.
Planning forward gave me four pages and no direction. Planning backward gave me a phone number.
Most plans fail because they start where you are.
Special shipment from Chatbase is heading out this week.
We can't show you what's inside yet.
But if you want one: comment below and we'll see what we can do. We kept a few aside.
when you're 20 you think you have all the time in the world
when you're 30 you realize you don't
when you're 40 you realize you never did
the time is now. it was always now. it will always be now
stop waiting for the right moment