A few weeks ago, we registered Alpha Bits as a company in Ho Chi Minh City. On paper it's a technology consultancy. In practice it's what we've been doing informally for years: building things with digital technology, learning as much as we can, and sharing what we find.
This post is our attempt to say what Alpha Bits is and what we think it could become. We don't have a five-year business plan or a pitch deck. We have a point of view and a balcony full of experiments.
What "Alpha Bits" means
The name means the alphas in digital bits.
We've always been drawn to digital technology, the raw bits of information that move through networks, run software, and increasingly shape how businesses and communities operate. "Alpha" is the ambition: to be early and good at this, and honest about what we know and don't know.
I took apart my first wireless router as a teenager, amazed that these tiny bits could travel thousands of miles through the air. The curiosity never went away. It just got more expensive.
Why now
We could have started this company three years ago or three years from now. There was nothing strategic about the timing. We had built up enough experience, enough projects and enough scars from real deployments that it felt dishonest to keep doing this on the side.
We're also watching a few shifts that we think will shape technology work over the next few years, and we want our bets on record.
AI will shrink teams, not replace them
The headlines say "AI will replace developers" or "AI will revolutionise business." Our bet is narrower and, we think, more disruptive: the tooling is getting close to the point where a small, skilled team, maybe even one person with the right tools, could do work that used to take fifteen or twenty people.
I've run ERP teams of 20-25 engineers, with technology leaders, designers, frontend, backend, mobile, QA and DevOps all on staff. My gut says this model has maybe three or four years left before something fundamentally changes. Today's AI coding assistants are crude, but they're improving fast enough to make me uncomfortable, in a good way.
If we're right, Alpha Bits is built for that world: a small team that punches far above its weight. If we're wrong, we'll still have built a good consultancy, so the bet costs us nothing.
Something interesting is happening in energy
On a completely different front, I've been running a thermal energy storage experiment on my Saigon apartment balcony since late last year, with sand, heat, sensors, Node-RED and a lot of patience. Storing energy as heat in ordinary sand sounds ridiculous. Maybe it is.
The data is interesting, though. The monitoring system we're building around it, with cheap IoT sensors and open-source tools on Raspberry Pis, is showing us something real. If the numbers hold, it could lead to patents, to products, or to a new direction for the company that has nothing to do with software consulting.
I don't want to oversell it. Most experiments fail. This one might sit on the balcony for years before anything comes of it, or go nowhere at all. Even so, rigging it with sensors, collecting data and analysing it methodically is teaching us things about IoT development that no textbook would.
Southeast Asia is underpriced
The talent here, especially in Vietnam, is remarkable relative to the global market. The problems are interesting too: manufacturers still running on spreadsheets, F&B chains with 200 outlets and no unified data layer, e-commerce operations held together with messaging apps and prayer.
Messy, real-world problems like these teach you things no tutorial will. We're betting that companies here will increasingly need people who can build systems from first principles, not just configure off-the-shelf products.
The infrastructure for digital nomads keeps getting better as well. I've written code in Hoi An cafe shops overlooking lantern-lit rivers, on balconies above rice paddies, and at tiny wooden tables squeezed between scooters. You can live well here and do serious work. That combination will bring more builders to this part of the world, and we want to be part of that community.
Open source has won (almost)
By "won" we mean that transparent systems are harder to compromise and easier to trust. Not everything is free. We've built serious infrastructure, including IoT monitoring, data pipelines and automation workflows, on open-source stacks that cost less per year than a single enterprise software license costs per month.
Our current HomeLab runs on about $2,100 in hardware and $12/year in networking costs. That covers Raspberry Pis, Orange Pis, a dedicated server, ZeroTier and Cloudflare, the whole thing. It hosts our own services and prototyping environments, and it may soon handle client workloads.
We think more businesses will figure this out. When they do, they'll need people who already know how these tools behave in production: people who have lived through the crashes, the config drift, and that one time a Node-RED flow melted wiring because of a missing safety character. Unfortunately, we are those people.
What kind of company this is
We're builders first, and the business exists to support that. It decides who we hire, which projects we take and how we spend our time.
Our people need hard problems. If a project doesn't teach us something new, it probably isn't the right project for us. That may sound idealistic, and maybe it is, but so far our most profitable work has also been our most technically challenging. We're betting that correlation holds.
We also want to be more than an agency. We want to organise workshops, run community events and share knowledge openly, and we see that as an obligation more than as marketing. The tech communities in Ho Chi Minh City and across Vietnam gave us our start, and the least we can do is give back.
We believe that in this new era code will become cheap enough to give away. If that happens, and we think it will, the value moves from writing code to knowing what to build and why. That's where we're investing.
What's on our desks right now
Rather than a list of services, here's what we're actually working on:
-
The thermal battery experiment: sand, heat, IoT sensors and data collection on the apartment balcony. It could become a patent, a product, or nothing. The data will decide that, not us.
-
Data infrastructure for a coffee chain: we're helping a multi-brand F&B operator in Vietnam unify its data across 200+ outlets. It has five different POS systems, and none of them talk to each other. The problem sounds boring until you're knee-deep in it.
-
HomeLab infrastructure: a distributed setup that handles everything from internal services to client prototyping. We'll document it in detail eventually. Everything we do runs on it, and it cost less than a decent used car.
-
PCB experiments: early tinkering with custom circuit boards for IoT applications. We have a hunch there's a product somewhere where education, IoT and robotics meet, but we can't see its shape yet. Hardware is hard; there's a reason nobody calls it "easiware." Some of these boards will end up in a drawer, and some might end up in something real.
None of these are marquee projects. They show us how technology behaves in the real world, which is messy, surprising and occasionally beautiful.
What we believe
We hold a few things to be true. We might be wrong about some of them, but they guide our decisions for now.
-
Tools change, curiosity doesn't. New platforms, frameworks and models will appear every quarter. Our instinct stays the same: be the first to try them, put up with the frustration of not knowing what we're doing, and come out the other side with working knowledge. Explore, struggle, master, share: that cycle is how we operate.
-
Build first, theorise later. We'd rather show you a working prototype than a presentation deck. If we haven't used something in a real project, we won't write about it as if we have.
-
Data first, opinions second. We've been developing a mindset we loosely call "Data-First Principle Thinking": raw data, properly collected and rigorously analysed, will lead you to conclusions that speculation alone never reaches. It applies as much to energy experiments as to software architecture and business strategy. Trust the numbers, document everything, and let the insights emerge.
-
Transparency is a feature. We publish our code when we can, because transparent systems are harder to break and easier to maintain. To us a GitHub link is proof that something works, more than a marketing asset.
-
Community is the long game. We don't know exactly what Alpha Bits will look like in five years. We do know that the relationships we build now, with other developers, with local tech communities and with people as curious as we are, will matter more than any client contract.
What's next
We'll keep building, and we'll write about it here: what works, what doesn't, and what we learn along the way.
Some of our experiments will fail publicly. Some will stall in prototyping for months or years, which is how invention goes. Some, we hope, will turn into something worth sharing. We won't pretend to know in advance which is which.
If you're a builder, a tinkerer, or someone who likes reading about how technology works in the real world, stick around. We don't promise polished thought leadership or quarterly trend reports, just honest stories from people doing the work.
If you're a business with a messy technology problem you don't know how to solve, we'd like to hear about it. We don't have all the answers, and working them out is the part we enjoy.
Alpha Bits. Builders first. Always curious.
This post was written during the first week of Alpha Bits' existence. We're a small team in Ho Chi Minh City with strong opinions about technology and no idea how this will turn out. Check back to see if we were right about anything.