We build the agent skills your services firm uses to deliver more client work — with the people you already have.
Live, no slides · Builders of Rudys.ai, co-built with SearchLab
Fixed price. Fixed scope. We don't sell hours of advice.
Free · 30 minutes
We join a call and run Claude Code against your actual work — your real report, brief, or dashboard. You see what's possible. No slides.
€5,000 · 1 day
We ship one focused agent skill, live, with your team in the room. You leave the day with a real artifact your team can use on Monday.
€15,000 · ~3 weeks
We generalise a full process into agent skills your team runs on real client work. We hand off. You own it. No ongoing dependency on us.
Need to bring partners or the broader team along before committing? Add a Stakeholder Day (€4,000) — a paid day where the team uses Claude Code on real work together.
Risk-free Build Day. We don't invoice the Build Day until we've agreed on the skill we're shipping that day. If you walk out without something usable, you don't pay.
"Most consultants give you a deck. Semicolon built the actual systems our team runs every day — and a SaaS product alongside it."
Concrete artifacts your team would run, not abstract "AI capabilities."
We don't have a skill for your work yet? That's fine — we generalise from your actual process. The Demo Call shows you what yours could look like.
Semicolon is led by senior AI engineers. Claude-native specialists, not consultants.
SearchLab was a Dutch Google Ads agency. Good at one thing, capped by team capacity. We worked together on two outputs from the same engagement: a complete agent-skill library their team uses to deliver client work, and a co-built productized SaaS (Rudys.ai) serving SMB customers their agency couldn't profitably take on.
Same team. Different leverage. They're now positioned as "AI as a Service" and Rudys.ai is live with paying customers (€19–€99/mo).
"Most consultants give you a deck. Semicolon built the actual systems our team runs every day — and a SaaS product alongside it."
A reusable, parameterised process that runs on Claude Code. Your team triggers it, it does the thinking — analysis, writing, recommendations — and produces a client-grade deliverable. A weekly client report, a project audit, a pitch brief. Custom-built for how your firm actually works, not a generic template.
The Demo Call is free (30 min). The Build Day is €5,000 — one day, one skill shipped with your team in the room. The Build Sprint is €15,000 — full process generalised into skills, ~3 weeks, handoff at the end. Repeat builds (skill #2, #3) cost less since we already know your stack and team.
No. Your team uses the skills, they don't build them. Running a skill is as simple as typing a command. We handle all the technical design and implementation. If you can use a chat interface, you can use what we ship.
One day, if you start with a Build Day — you walk out with a working skill. The Build Sprint ships the full process within ~3 weeks. There's no "AI strategy phase" before we build.
You own the skills. Your team runs them. We don't sell maintenance retainers — there's nothing to maintain. Most clients come back when they want the next process built, not because anything broke.
Consultants give you a strategy and leave. We build the actual systems your team runs every day. The proof is Rudys.ai — a live SaaS we co-built with SearchLab, with paying customers right now. We don't talk about what AI could do. We ship it.
Drop your details and we'll send a calendar link within 24h. On the call we run Claude Code on your actual work — live, no slides. If the Build Day's a fit, we ship a skill within a week. You don't pay if you don't walk out with something usable.