4 min read

The flywheel turns: from services to products

studioproductsannouncement

From the beginning, we said it plainly on our own homepage: recurring problems become products; services sharpen the blade.

Today the flywheel completes its turn. Slicing Water is now an AI-first product studio. We are no longer taking on services engagements. Everything we learned doing hands-on systems work — every legacy pipeline we untangled, every chaotic document flow we tamed, every manual verification step we watched a team suffer through — now goes into products we build and ship ourselves.

Why the turn

The services era did exactly what it was designed to do. It put us inside real operations, elbow-deep in real friction. You cannot build a product for a problem you haven’t touched, and we touched a lot of them.

What we found is that the frictions repeat. The shapes change — a spreadsheet here, a PDF queue there — but the core drag is the same across companies, industries, and stacks. When the same wound keeps appearing, you stop treating patients one at a time and start making the medicine.

That is what a product is: the same cut, made once, that holds for everyone.

What a blade looks like

Our products are micro-SaaS, and that word “micro” is doing real work. Each product is deliberately narrow: it finds one specific, recurring friction and removes it completely. No platforms, no suites, no fifteen-tab dashboards. A blade does one thing — it cuts.

And they are AI-native by default. Agents, automation, and model integration are the substrate, not a feature checkbox. Agentic automation is also how we build: agent-driven development inside deterministic guardrails lets a small studio ship at the pace of a much larger team, without the sprawl.

What stays the same

The name, the tagline, and the principles all carry over — because they were never really about consulting. They were about making cuts that hold:

  1. The cut holds. Deterministic boundaries around every stochastic part. If the AI wobbles, the code catches it before it touches your data.
  2. No dependency tax. We hated seat-license bloat when we were removing it for clients, and we won’t ship it ourselves. Simple pricing, small surface area, no lock-in theatrics.
  3. Operator empowerment. Our products remove the boring parts of the job, and leave people in charge of the judgment calls.

Friction is like weeds. Trim the leaves and it grows back. Slice it at the root and it doesn’t. That was true when we did it by hand; it’s truer now that we can ship the blade.

What’s next

The first products are on the whetstone — in active development now. We’ll document the build here in the field notes: architectures, patterns, agentic pipelines, and the decisions behind them, as we go.

If you want the first cut, join the waitlist. One email when a blade ships. Nothing else.

And if you live with a friction you want dead, tell us about it. The best blades start as someone’s daily annoyance.

friction doesn’t grow back.