TikTok Orders Took 3 Weeks to Arrive. He Built the Fix.
Why Richard Chen left Big Tech to start FlexHaul.
What do you do when your TikTok order takes three weeks to arrive?
If you’re Richard Chen, you quit your job and build a logistics platform from scratch.
That frustration became the spark behind FlexHaul-an AI-powered operating system for warehouses and small trucking companies, many of which still run on Excel, phone calls and hope.
In this episode of The Founders Chat, Richard shares how he went from shipping ML models to shipping freight-and why logistics might be the most overlooked startup opportunity of this decade.
It started with one slow delivery
TikTok was betting big on live shopping.
But when Richard used it himself, his package took three weeks to arrive. By the time it landed, he’d forgotten he’d even ordered it.
“The problem wasn’t content or demand. It was logistics.”
Amazon set the bar. Everyone else fell short.
That became FlexHaul’s origin insight.
The real opportunity wasn’t in e-commerce. It was in warehouses.
Richard started with integrations-FedEx, UPS, USPS-aiming to help e-commerce sellers route smarter.
But something unexpected happened.
Warehouses showed more urgency than storefronts.
They weren’t managing 1 brand. They were managing dozens.
Orders in spreadsheets
Pricing by guesswork
Fulfillment by phone call
FlexHaul pivoted to serve them directly.
Today, they focus on:
Warehouse operators
Mid-market logistics teams
Small trucking companies (90% in the US own fewer than six trucks)
“More than 25 percent of US trucking companies don’t use any software.”
Why others failed where FlexHaul is winning
Richard studied logistics startups that collapsed-Convoy, Zum.
Their mistake? Overreliance on AI models that couldn’t adapt to economic shocks.
When COVID hit, demand spiked. After, it cratered. Their algorithms couldn’t keep up.
“Human brokers adjusted. AI didn’t.”
FlexHaul took a different path:
Use AI as leverage. Keep humans in the loop.
The engineer-to-founder transformation
Richard spent over a decade building at Meta and TikTok.
He loved building-but realized he didn’t know how to sell.
“As engineers, we think: if I build it, they’ll come. That’s not how it works.”
He had to learn:
Sales
Contracts
Pricing
Customer discovery
Leadership and delegation
His edge? He could offer to build custom integrations for customers-himself.
“That closed deals. Fast.”
How FlexHaul uses AI
The stack:
Python FastAPI
React Native
GCP + Docker
Cursor
They use ChatGPT-5 for product planning.
Then run PRDs through Cursor:
Reads the codebase
Connects to GCP
Executes shell commands
Flags bugs
Suggests fixes
Plans projects
“Cursor is our second engineer. We fix bugs before customers notice.”
Why he builds solo
Richard had co-founders before. It didn’t end well.
“I build faster alone. Fewer misalignments. Zero drama.”
He’ll add a co-founder later-once the foundation is solid.
The weight of it all
Richard came to the US as an immigrant. His dad worked overseas. Responsibility hit early.
Now:
He works more than ever
He takes melatonin to sleep
His brain won’t switch off
“I’ve learned more in 12 months of founding than in 15 years of employment.”
The mistake that nearly broke him
Early on, he built first. Sold later.
Warehouses destroyed his assumptions.
“They showed me how little I knew. That’s when I leveled up.”
What new founders get wrong with AI
AI is fast. But that’s the danger.
“If you don’t have judgment, AI just helps you ship the wrong thing faster.”
Engineering still requires:
Debugging
Architecture
Systems thinking
Scar tissue from 3 a.m. outages
Success, redefined
“Success is owning my time. That’s what I’m building toward.”
Message to customers
FlexHaul solves problems fast. Sometimes before you even see them.
They’re not selling features. They’re building reliability.
Message to investors
FlexHaul is bootstrapped.
“We’ll raise once the revenue base is stronger. I want to de-risk this before I bring in capital.”
His superpower?
“I never quit. And I always assume I might be wrong. That lets me learn faster than everyone else.”
Full Episode below
What’s your story? Leave it in the comments!
The transition from engineer to founder is such a critical shift. Building the product is one thing, but learning to sell, price, and discover actual customer neds is completely diferent. The ability to offer custom integrations as a closing strategy is clever use of technical skillset.