Everything here is something we've built and still maintain. If your problem doesn't fit one of these, we'll tell you.
Make people do less repetitive shit.
Give the machine the work it does better than you. We connect the tools you already pay for, put AI where it earns its place, and remove the parts of the job nobody should be doing by hand.
The problem
Orders get typed over from email into the system. Invoices get matched by eye. Someone answers the same twelve questions every week. It works until the person doing it is ill, busy or gone.
What we build
When it makes sense
You can point at a task and say "this happens every day and it's always the same steps." If you can describe the rule, we can usually automate it. If you can't, we'll help you find out whether AI can learn it.
Examples
Expected outcome
Hours back every week, fewer errors, and nobody's job resting on remembering.
01 · 1 week
Audit
We sit with the people doing the work, map the repetitive tasks, and pick the ones worth automating first.
02 · 2–3 weeks
Pilot
One workflow, automated end to end, running next to the manual process so you can compare.
03 · 2–4 weeks
Roll-out
Extend to the rest. Human approval stays where it matters; the machine handles the rest.
04 · Ongoing
Monitor
It runs, we watch. You get a monthly note on what it did and what it saved.
Typically 4–8 weeks to the first automation running in production.
Turn an idea into something people can use.
An idea isn't useful until somebody can use it. We take it from "we should really have a system for this" to a product with real users — architecture, interface, deployment, and the part that gets skipped: shipping.
The problem
The business runs on five tools and a spreadsheet that only one person understands. Off-the-shelf software does 70% of what you need and fights you on the other 30%. Or you have a product idea and no team to build it.
What we build
When it makes sense
You've outgrown the tools, or the process is the business and no template fits. Or the idea is validated enough that the next step is a working version people can touch, not another deck.
Examples
Expected outcome
One system instead of five, and a product with real users on it.
01 · 1–2 weeks
Discovery
Process mapping, data model, fixed scope. You know what you’re getting and what it costs.
02 · 4–8 weeks
MVP
Weekly demos on real data. The first version people put work into.
03 · 1–2 weeks
Launch
Migration, training, go-live. We’re in the room (or the call) when it goes on.
04 · Every 2 weeks
Iterate
Releases on a rhythm. What users ask for, built; what they don’t use, removed.
Typically 8–12 weeks to a product your team is actually using.
Build a website people remember.
Most sites look like the template they came from. We design and build the ones that work as hard as the business standing behind them. Fast, findable, and with something to say.
The problem
The site was built years ago, by someone who's gone, on a platform nobody dares touch. It's slow, it looks like every competitor, and nothing on it says why anyone should pick you.
What we build
When it makes sense
Your work is better than your website. You're rebranding, launching, or tired of apologising for the site in meetings. Or you have a complicated thing to explain and a page is the right place to do it.
Examples
Expected outcome
A site you'd send someone to on purpose, and one you can edit yourself.
01 · 1–2 weeks
Strategy & copy
What the site has to say and do. We write the words before we draw the boxes.
02 · 2 weeks
Design
Real content, real motion, one direction. You see the page, not three mood boards.
03 · 2–3 weeks
Build
CMS, performance, structured data and accessibility built in from the first commit.
04 · 1 week + ongoing
Launch & measure
Go live, then watch what people actually do. Adjust.
Typically 6–8 weeks from first call to launch.
Make the internet understand you.
Search changed. AI assistants now answer the questions that used to route to your website. We build the structure that gets you found by search engines and quoted by answer engines — the same work, done properly once.
The problem
You rank for your own name and nothing else. When someone asks ChatGPT or Google's AI "who does X in Y", you're not in the answer — because your site never says clearly what you do, for whom, and why.
What we build
When it makes sense
You're building or rebuilding a site anyway — do it right from the start. Or the site is fine but invisible, and you'd rather be found than run ads forever.
Examples
Expected outcome
Found for what you do, by people and by the machines they ask.
01 · 1 week
Audit
Technical health, content gaps — and what AI assistants currently say about you.
02 · 2–3 weeks
Foundations
Structure, speed, schema. The things machines need before they’ll trust the words.
03 · 4–8 weeks
Content
Service and location pages with real substance, written to be quoted accurately.
04 · Monthly
Measure
Rankings, citations in AI answers, and the one that matters: leads.
Results compound over 3–6 months. Anyone promising faster is selling ads.
Make your app work for more businesses.
You built an app that runs one business well, maybe vibe-coded over a few weekends. We turn it into a product other businesses can sign up for: their own data, their own settings, one version to keep up to date.
Mike leads this work. He is our back-end co-founder and builds systems that scale.
The problem
The app works. The database is connected, payments go through, customers get their messages. But your business is written into the code, one login can see everything, and the payment keys sit in a file. A second business can't just move in.
What we build
When it makes sense
Another business has seen your app and asked for it. Or you can name five that work like you do. If a second business can't run on it without code changes, it isn't a product yet. That's the gap we close.
Examples
Expected outcome
One app, many businesses, and one update that reaches all of them.
01 · 1 week
Audit
We read the code and the data, and check for leaked keys and missing login checks.
02 · 2–3 weeks
Split
What every business needs goes into the core. What only yours needs becomes a setting.
03 · 3–6 weeks
Tenants & connectors
Each business gets its own walled-off data. Payments, accounting and messaging plug in per customer.
04 · 2–4 weeks
Pilot
A second business runs on it. Every code change it needs is a gap we close.
Typically 8–14 weeks from your app to a second business running on it.
Most projects touch two or three of these. Thirty minutes is enough to tell you which, and whether we're the right people to do it.
Book 30 minutes → Or throw us the problem you think can’t be solved →Read the questions people ask us first →