About Me

Almost every person I talk to asks, "Why freight?" I get it, it’s not a clear-cut choice for a software engineer like me to take on such an arcane industry. While people can sense my enthusiasm and energy (I hope), I’ve always struggled to clearly articulate why. The truth is: there’s a ton going on in my ADHD brain.

It’s like trying to dump my mega-long context into ChatGPT’s 128k token window and getting a “500 error.” Then, I switch to Gemini’s 2M token window, and it times out. I feel like I have to step back and perform multi-step prompting or chain-of-thought reasoning to semantically embed and retrieve the important needle in a haystack. By the time I’m ready to share something coherent — 15 minutes have passed and the topic has already moved on.

In a nutshell, my reasons boil down to:

1. An immense potential to change people’s lives

2. Unprecedented “why now” moments in freight, with a laundry list of contrarian hypotheses and experiments to prove

3. A personal relatedness

4. A top-notch team

1. Immense impact

• If Stripe’s growing the GDP of the internet, a great company in freight would become the backbone of that growth, driving global trade Forward (no pun intended), digitally and physically.

• A whitespace $200B freight industry, largely due to an elephant in the room: resistance to change, which is just human nature. Freight brokers and forwarders still favor mano-mano (”pen and paper” in Filipino) and laborious Excel wizards, despite the rise of software and AI.

• Mission-critical (= also means a higher risk of messing something up):

  • If LinkedIn’s search feature is inaccurate, you can just retry.

  • If you’re egregiously disbursing $1M instead of $100k, it’s disastrous, but at least you can track and trace the money, issue refunds / reverse the transactions.

  • But if you ship something wrong from Kansas to Busan, the cost is irreversible. A digital mistake has already turned into a real-world problem. That’s why fast, accurate, and cost-effective solutions are critical to avoid that kind of catastrophic damage.

2. Unprecedented “why now” & contrarian hypotheses

At least 3 tailwinds:

1. Economic Impetus

   • Historic recession in freight → cost cutting pressure towards freight companies → buying incentives. Volumes dropping, rates falling, one of the 1st Christmas without peak season.

   • Perfect storm of contraction: Freight juggernauts (Convoy and Flexport) are struggling to consolidate fragmented market. This paves the way for for next-gen freight tech companies to land grab brokers and forwarders (and even shippers) in a market that remains predictably fragmented.

   • Potential dream to become a single source of truth for freight. Mastercard of freight.

2. Cultural Buy-In & Behavior Shift

   • There’s a major shift as companies prioritize maximizing revenue per employee (top-line growth) over merely cutting costs (bottom-line savings).

   • The era of second-generation freight owners: creating a new freight buyer archetype, digital native

   • Market change. Of the $200B freight industry, <5% is IT budget, up to 40% is labor budget (AI would act as a labor/service instead of IT)

3. Technology Advancement

   • Ease of adoption: LLM and GenAI are changing the game from "Software as a Service (SaaS 2.0)" to "Service as Software (SaaS 3.0)" . Selling work, not software. 0 change - keep your Excel, email, and phone calls like how you interacted with your employees.

   • Affordability: Turning labor cost from fixed (human/BPO) to variable (computing). Making it scale with revenue. When volumes drop, labor costs drop too, keeping margins intact.

   • Repeated: Train once and for all. 24/7. Machines don’t churn

Contrarian Hypotheses to Prove

1. Building human-powered AI, not AI-powered humans.

   • The team learned how to do this at a billion-dollar YC-backed fintech (think early days of Plaid/Stripe)

     • they launched semi-manual lightweight APIs to accelerate the feedback loop while slowly developing a proper backend infrastructure.

   • Now, they plan to do the same in freight:

     • Deploy AI employees to handle end-to-end tasks.

     • Begin with a self-learning system enhanced by domain-expert supervision

     • Gradually increase automation, aiming for 99% self-learn software and 1% human involvement (or even 0%).

   • The team has both experience and service cost arbitrage (props to their global talent coverage)

2. Tackling human-like front-office roles, not just back-office work, to coexist and scale alongside humans (without reducing headcount)

3. [The list goes on …]

3. Personal Relatedness

There's a saying: your career is the accumulation of all your life experiences, and I think mine are:

1. Building “infrastructure”

2. Everlasting curiosity to learn, from building Lego blocks to biology experiments, to international and immigrant experiences.

3. Rediscovering my roots

4. Past NGO stint: Future of work and economic empowerment

Building “infrastructure”

• Saw a parallel between freight and payment infrastructure after working in fintech for a while:

  • Payments: moving money; freight: moving data and goods — inevitable to build something insanely great: accurate, fast, cost-efficient, and, not to forget, human-like.

  • If payments are not highly digitized, freight is currently poorly digitized.

Everlasting curiosity to learn

• new domain, freight, hungry for knowledge

• Intellectual honesty to seek truth: always love experiments, validating (+ killing) hypotheses, even if the ultimate answer is “no”, you can accumulate the learnings to shoot for a more accurate outcome

• Freight is a global play: there might be discrepancies in different markets to be addressed. Taste is required, beyond simple consolidation, per usual Valley playbook. Plus, I love learning about the world, one step at a time.

Rediscovering my roots

• Born into a family of truck drivers, early exposure always made me curious about the lives of my predecessors. going back to roots.

Past NGO stint

• Always curious about how humans and AI can coexist, how we can help people do more meaningful work driven by their passion (which theoretically should lead to greater economic opportunities) instead of getting stuck with mundane tasks

4. Top-notch team

Scrappy but solid

• Early days, the team flew one-way to wherever customers are to build in-person, from warehouse, airport, and even CU (씨유) mini-markets. From Singapore to Manila, SF to Busan, the Netherlands to Indonesia and the rest of North America.

• “Technically” a dropout Silicon Valley engineer, a Penn/Stanford product expert, and a UoFT alum who was the 1st business hire at an acquired AI startup (University of Toronto, same uni as OpenAI cofounder Ilya Sutskever and AI pioneer Geoffrey Hinton), bringing Valley experience and a contrarian product playbook from Asia and Latin America

• Mafia of early employees from 3 YC-backed companies from zero to billion-dollar. Full coverage across dev, product, sales.

• Other team members come from unique backgrounds, geophysics, aerospace, and even music engineering, who boldly made the leap into software development, hungry for knowledge and learning with the tenacity of a honey badger

At the risk of sounding hyperbolic, I’d say that I’m willing to spend the next decade solving this problem, and ideally, I want this to be my last job ever.