Every question we get asked, answered straight.

Most builds ship a working first version in two to six weeks, depending on the job. We work in two-week sprints, so you see progress every fortnight rather than waiting months for a reveal. The mapping phase before the build tells you the timeline for your specific job before you commit.

We spend time inside how your team actually works, watching where the week goes. You get a written map of the jobs AI could genuinely take on, what each would save, and a plan with a price. If the honest answer is "don't build anything", that's what the map says.

Yes, that's the default. We build around the tools you already pay for, whether that's your CRM, your accounts package or a stack of spreadsheets, rather than asking you to replace anything. If a system has no way in, we'll say so at the mapping stage, not after.

Messy data is normal and rarely a dealbreaker. Most businesses run on spreadsheets, inboxes and half-tidy systems, and we build with that in mind. The mapping phase looks at what you actually have, and if the data genuinely can't support a build, we tell you before you spend on one.

No. We build systems your team can run without writing code, and the optional monthly support covers the technical upkeep. If you do have developers, co-ownership means they can see inside everything and take over whenever you like. Either way, day-to-day use needs no technical staff.

You tell us and we fix it. Systems on a monthly support plan are monitored, so we usually know before you do. Because every build keeps a person in charge of the output, a fault means work slows down for a day, not that wrong answers go out the door.

Yes, any part, at any time. Every system we build has a plain off switch and degrades politely: the work goes back to being done the old way rather than piling up unseen. Nothing we build is load-bearing in a way you can't undo.

Sometimes, where it's the right tool. We build with whichever model suits the job, cost and data sensitivity, and we're not tied to any vendor. What you're paying for isn't access to a model; it's the system around it that fits your business and keeps a person in charge.

An agency builds what you specify; we help you work out what's worth building first. Every engagement starts with a paid mapping phase inside your business, the builds are AI systems rather than general software, and you deal directly with the two founders who do the work.

Yes. We're based in London and most of the work happens remotely anyway, over calls and inside your systems. For UK clients we'll come to you when it helps, especially during the mapping phase. Distance has never decided whether an engagement works.

Every two weeks you see working software and decide what happens next. Each sprint has an agreed goal, ends with a demo of the real system on your real work, and you can pause or stop at any sprint boundary. No six-month build with the results hidden until the end.

It means the system is yours as much as ours. You can see the code, the prompts and the data flows, run it where you choose, and keep it if we part ways. We keep the right to reuse general techniques, never your data or your business logic.

A little, and we build it into the handover. Most systems we ship look like a report arriving, a queue getting shorter or a checklist to review, so the training is usually an hour or two, not a course. If a tool needs a manual to use, we've built it wrong.

The repetitive ones with clear rules and a painful monthly cost. One report we're replacing eats a team eight to ten days every single month; that's the shape to look for. The Impact Call exists to find that job in your business, and the mapping phase confirms it.

Anything where judgement is the job: hiring decisions, client relationships, pricing calls, or anything you'd need to defend later without being able to explain it. We automate the assembling, checking and chasing around those decisions, and keep a person in charge of every output that matters.

The two founders. Tristan does the thinking with you: scoping, priorities and the plan. Jim, our CTO, builds it, bringing 25 years in tech and three exits: Just Eat (IPO), Time Out (IPO) and @Leisure (sold to OYO). No delivery team you've never met, no handoffs.

Mostly UK small and mid-sized businesses: big enough that someone loses days each month to repetitive work, small enough that there's no internal team to build the fix. If you're a founder or director who can point at the job that hurts, you're the right size.

Nothing formal. Come with a rough sense of where your team's week goes and which jobs feel heaviest; a guess is fine. No documents, no data, no slides. The call does the rest, and you leave with a one-page opportunity map whether or not we go further.