Ronald Schlenker · PureState Let's talk
Training · Consulting · Engineering

Trying to put AI in your product and finding it harder than the keynotes suggested?

Leaders want AI working by next Tuesday. Most teams need to get three foundations right first — and none of them are on the conference slides. The bridge from "AI is capable" to "AI delivers measurable value in your product" is engineering, not a platform.

Three foundations first.

  1. A problem in your product that AI is actually well-suited to solve.
  2. A codebase and data story your team can iterate on — not a demo.
  3. Evaluation that tells you whether it works, before it ships.
Recognised

“Recognised F# Expert” — F# Software Foundation, Applied F# 2019.

Fifteen years in .NET · invited to fsprojects by Don Syme.

Services

Training Workshops · talks
Consulting This page
Engineering AI · DSL · tooling

How the AI Adoption Review works.

01 / Engagement

Who this is for.

02 / Fit

Book a call if you…

  • own a real product and need AI to do a real job in it.
  • have a team you want to keep, not replace.
  • want a senior opinion that isn't selling you a platform.
  • prefer a written, defensible argument over a deck.

Skip this if you…

  • want someone to manage a fifty-person offshore team.
  • are looking for an inspirational workshop and a certificate.
  • need staff augmentation by the seat.
  • expect AI to be working by next Tuesday.

Things already built and shipped.

03 / Proof
Recognition

F# Software Foundation

Recognised F# Expert — Applied F# 2019. Recurring features in F# Weekly.

foundation.fsharp.org →
Open source

FsHttp · fsprojects

Invited by Don Syme (creator of F#) into the official fsprojects org. 499★, 128 dependents.

github.com/fsprojects/FsHttp →
Teaching

BobKonf 2024 speaker

Computation Expressions in F# — full tutorial track with David Schaefer.

bobkonf.de →
Community

Amplifying F#

Co-host on the community format run with G-Research OSS. Long-form practitioner conversations.

github.com/SchlenkR →

If you're curious who's behind this page.

Outside client work, I build. Here's the short version — what this guy actually does when he's not consulting.

  • PXL Clock — a 24×24 programmable LED display I co-founded with Sefa, engineered end-to-end by two people, shipping in limited batches from Frankfurt.
    pxlclock.com →
  • TypeFighter — my experimental programming language. A modern, inference-first type system where records match by shape, not by declared name. Research-grade, explained end-to-end.
    github.com/SchlenkR/TypeFighter →
  • And plenty more on GitHub — FsHttp, Trulla (type-safe templates), TheBlunt (parser combinators), LocSta (stateful stream processing).

Common objections.

04 / FAQ
Sounds expensive.

Less than hiring a senior full-time. Scoped explicitly — no retainer trap, no seat-based staffing arithmetic. You know the boundary before we start.

We already have an agency — what's different?

An agency builds more capacity. I make your existing team capable. Different job. The deliverable is a document and a defensible decision, not a delivery team.

Can we do this remotely?

Yes. DACH and remote EU, both supported. Working sessions over video, code walkthroughs in your repo, written output you can forward in the next board meeting. Onsite possible for team-facing days.

Start a conversation about your AI adoption.

Tell me where you are, what you've tried, and what you're hoping this unlocks. You'll get a reply, not a proposal deck.

Email hello@schlenkr.dev Subject: AI Adoption Review inquiry