Skylab · Build & maintain
Running a small business already takes everything you have. Building and looking after the software behind it shouldn’t be another job you have to learn.
Skylab is a small, senior team that designs, builds, and maintains the systems your business runs on — sized and priced for a business that isn’t in the software business itself. We start with a two-week Discovery Sprint, not a six-figure proposal.
The problem
A lot of small businesses run on the same three things: an owner or manager who remembers how everything actually works, a spreadsheet nobody else fully understands, and whatever software got bought years ago and never revisited.
None of that is unusual — it’s how most small businesses run, because until recently, building something better wasn’t worth the time, and usually wasn’t something you could afford.
We re-enter the same information into three or four different systems, by hand, more or less every day.
The person who set up our process left a couple of years ago. Nobody's fully sure why it works the way it does, so nobody wants to touch it.
We tried an off-the-shelf tool. It covered maybe sixty percent of what we needed, so now we run it alongside the spreadsheet instead of using it on its own.
I know we're losing money to things falling through the cracks. I don't have the time or the people to go find out how much.
We got a quote from a software firm once — six figures and the better part of a year, for a business our size. We said no and went back to what we had.
Everyone here already does three jobs. ‘Fix our systems’ has never been anyone's actual job.
None of this is because the people running these businesses are careless — it’s because their time and their budget both go to the business itself, the thing that actually makes money, and there’s rarely anything left over for the technology behind it. The pattern shows up almost everywhere: manual work that eats real hours every week, a process that depends on one or two people remembering how it’s supposed to go, and a growing list of small breakages nobody has time to fix properly. Off-the-shelf software rarely closes the gap, because it’s built for a generic business rather than yours. And the firms that could build something custom are usually priced and staffed for companies much larger than yours.
A case in point
A major nonprofit organization: advising on, and building, their biggest technology modernization in twenty years.
Skylab worked with the organization at a technology advisory level — from defining what needed to be built, through architecting and delivering it, to supporting and extending it in production today.
- 20 yrsOn MS Access & spreadsheets, before the rebuild
- ~10 yrsThe rebuilt system has run in production
- OngoingStill advising, building, and supporting it today
The client, a large nonprofit organization, had run core parts of their operations on MS Access databases and spreadsheets for roughly two decades. When they decided to modernize, Skylab came in at the technology advisory level — not just to write code, but to help figure out what should actually be built. Gayan led the discovery work and wrote the RFP that defined the project, then worked with the team through the architecture and build of a custom system to replace the old Access-and-spreadsheet stack. That system went to production, and it’s still the system the organization runs on today, about ten years later. Skylab is still involved: not just keeping it running, but building new features and extending it as the organization’s needs change. It’s the largest technology modernization the organization has taken on in the last twenty years, and Skylab has been their technology partner across the whole arc of it, from the first discovery conversation to whatever gets built next.
The honest part: this project predates AI-assisted development. AI did not write this system, and we won’t claim it did. What it does show is the pattern Skylab is built around — understanding a business well enough to know what to build, building it properly, and staying on as the people who keep it working and growing over time. What’s changed since then is that AI now makes that same kind of work faster and more affordable to deliver, which is part of why Skylab can offer it to businesses that could never have justified a project on the scale of that original build.
How we work together
One ladder, three steps. Each one is priced and scoped before you commit to it, and each one de-risks the next.
Fixed pricing, shown here — you shouldn’t have to book a call to find out what something costs.
Discovery Sprint
$5K–$10K
The front door. We spend one to two weeks learning how your business actually runs, then hand you a ranked list of where AI-based systems would help most, plus a concrete plan for fixing the first and biggest one. You keep the operating map either way — whether or not you go on to hire us for the build.
First Build
$25K–$60K
Your first working system, priced and scoped up front. What it covers depends on where your business bleeds most — often it's turning manual data entry, confirmations, and follow-up into something that runs with far less hands-on work, so your team spends its time on what only people can do. Whatever process you're running today keeps going, untouched, until the new system has earned its place beside it. The price is a fraction of what a $150K–$1M enterprise system would run, because delivery is AI-native and the team is small.
Own & Extend
$1.5K–$4K
We host what we build, keep it running, and keep extending it as your needs change — the next feature, the next workflow. It's month-to-month, and you own the code the whole way through, so you can leave anytime and take the system with you. Building something once is a project. Keeping it running and growing is where most of the value actually shows up.
The first two rungs are priced against the time you’re currently losing to manual work, not against some abstract IT budget. Free up real hours in a week, and most owners find they can take on more business, serve customers better, or just stop losing sleep over the thing that breaks if the wrong person is out for a week.
You own the code. Always.
How the options compare
Most of the options available to a small business fall short in one of two ways: they can’t actually build real software, or they won’t stay involved once it ships.
Skylab is built to do both — figure out what to build, ship it, and stay on as the people who keep it running.
For most small businesses, the real alternative to hiring anyone isn’t another consultant — it’s continuing to run on what you already have. That’s a reasonable choice, right up until the spreadsheet breaks, or the one or two people who understand the process aren’t available, and there’s no backup plan. Once it’s worth looking at other options, here’s roughly how they compare:
- Keep running on what you have todayIt works fine until it doesn't, and when it breaks there's usually no backup plan in place.
- Hire someone in-houseYou gain a person, but the knowledge is still stuck in one head. It's just a different one now, and you've taken on a manager's job as well.
- Off-the-shelf CRM / ERP / industry softwareBuilt for a generic business and usually priced for a much bigger one. You still end up doing a lot of manual data entry, and you're bending your process to fit the tool rather than the other way around.
- AI or technology strategy consultantsYou get a report and a set of recommendations. Whether any of it gets built is a separate project — one you still have to find someone else to run.
- Automation or no-code agenciesThey can wire together a quick automation, but it tends to be brittle, and there's usually nobody who can build the more substantial system underneath it if you actually need one.
- A large software firm or systems integratorCapable, but priced and structured for enterprise clients — six-figure budgets and multi-month timelines that rarely make sense for a business your size.
- An independent freelance developerCan be a good fit for something small and self-contained. For anything that needs real discovery up front and support after it ships, freelance engagements are often just a build with no plan for what happens next.
Why the structure works this way
The engagement is structured so you’re never risking more than the step you’re on.
The Discovery Sprint proves we understand your business, for a fixed, modest cost. The First Build proves the system works, at a fixed price, with your existing process still running underneath it. Only once that’s proven does the ongoing relationship start — and even then, you can leave anytime with the code in hand.
- “Skylab is a small team — what happens if something happens to you?”
- You own the code from day one. It lives in your repository, not ours, so there's no lock-in regardless of what happens to Skylab down the line. And the ladder means you're testing the relationship with a small, fixed engagement before you ever depend on us for something critical.
- “My whole business runs on what I have now. I can't afford for it to break.”
- That's exactly why we don't start with a rebuild. We start with discovery, and when we do build something new, it runs alongside your existing process until it's proven itself. You're never forced into a single cutover point where everything has to work perfectly on day one.
- “I've paid for a project before that never really shipped, or fell apart once the person who built it moved on.”
- That's the most common failure mode in this kind of work, and it's the reason the ongoing relationship is built into the model rather than treated as an afterthought. A long-time nonprofit client's system has been in production for about ten years, and Skylab is still the team supporting and extending it — that's the standard we hold ourselves to.
- “How do I know you can actually deliver this?”
- The clearest answer is the work with a long-standing nonprofit client — an advisory and delivery relationship that's lasted roughly a decade, on what became the organization's largest technology modernization in twenty years. Skylab has also done work with BCAA and Nike. That's alongside twenty years in software and a computer science and AI background from UBC.
- “This is real money for a business our size.”
- The ladder is priced so the first two steps are proportionate to the size of decision they represent — the Discovery Sprint is a small, bounded cost to find out if this is worth pursuing, and the First Build is fixed and scoped before you commit to it. Most owners find the return shows up in hours freed up and mistakes avoided, more than in any single number.
- “What if we go through the Discovery Sprint and it turns out this isn't worth pursuing?”
- Then you still walk away with a clear map of how AI-based systems could help your business, and no obligation to continue. It's designed to have value on its own, independent of whether we ever build anything together.
Selected workBCAA · Nike— across more than one client, and more than one industry.
Insights
What we’re thinking about.
Short, practical notes on where AI-based systems actually help a small business, and where they don’t — from the LinkedIn feed and the Skylab journal.
Import AI 464: Fable writes GPU kernels; AI automation; and analog computation
Most weeks in AI produce a lot of noise and very little signal. Every so often a week produces two numbers that, read together, are hard to shrug off.…
The summary is where your agent forgets
Most conversations about AI agents fixate on intelligence. Can it reason, can it plan, can it write code that runs. Those are the right questions…
Using Local Coding Agents
For roughly two years, the default answer to "which AI should write code with us" has had a single shape. You picked a subscription. You paid a frontier…
Get started
Book a Discovery call.
Two weeks, a fixed price, and a clear map of your business and where AI-based systems could help — yours to keep whether or not we end up working together beyond that.
Tell us a bit about how your business runs today, and what feels most fragile or manual about it. We’ll get back to you with a straight read on whether a Discovery Sprint makes sense — not a deck, not a sales pitch, just an honest answer.
Prefer email? hello@skylabconsulting.com
Questions
Straight answers.
- How long until something actually works?
- The Discovery Sprint is 1–2 weeks. Your first working build typically ships in weeks, not the quarters a large enterprise project would take, because delivery is AI-native and the team is small.
- Do I own what you build?
- Yes, always, at every rung. It lives in your repository. There's no lock-in — you can leave the month-to-month relationship anytime and keep the system.
- What happens to my current process during the build?
- It keeps running, untouched. We run the new system alongside the old one until it's proven itself, rather than switching everything over at once. There's no single point where everything has to work perfectly on day one.
- What do you need from me to start?
- A conversation. The Discovery Sprint maps how your business actually runs — you don't need to prepare anything or know what to build. That's our job.