01

Ronald Schlenker.
Independent practice for software teams that take engineering seriously.

PureState IT Consulting  /  Training · Consulting · Engineering

I help engineering teams design, build and operate software that holds up under real load — and I teach them to do it themselves. Current focus: applied generative AI, agents and LLM engineering, grounded in two decades of functional programming, .NET and domain modelling. No buzzwords. Concrete projects, measurable progress, deliberate craft.

02
a / Training

Workshops and in-depth courses for engineering teams.

Functional programming in F# and C#, domain modelling, generative AI for developers, LLM engineering, agent architectures. On site, remote, as multi-day deep dives or focused one-day sessions. Materials and exercises included.

b / Consulting

Architectural review, technical strategy, and hands-on pair work.

I join your team for a limited engagement: diagnose the system, clarify the problem, recommend the next three moves, and write the code alongside you. Short feedback loops, written deliverables, no ivory tower.

c / Engineering

Build with me, or have me build it.

Prototype to production: AI-assisted tooling, data pipelines, domain-specific languages, developer platforms, embedded experiences. Shipped through .NET, F#, TypeScript — or whatever the problem actually requires.

03
No.
Project
Note
001
Vide
A functional reactive UI library for F#. Incremental composition, explicit state, no VDOM.
002
FsHttp
A pragmatic, composable F# DSL for HTTP. Widely used across the F# community.
003
FluX
A computation-expression-based approach to real-time audio synthesis in F#.
004
LocSta
Local-state combinators — a compact functional model for stateful streams.
005
Trulla
Typed text templating for .NET — compile-time safety for generated content.
006
PXL Clock
A programmable 24×24 RGB pixel display. Side project via Cumin & Potato GmbH — hardware, firmware, browser simulator, SDK.
04
01
Types are a tool for thinking, not a tax. Model the domain first; the code follows.
02
Generative AI earns its keep when it is engineered like any other system — with boundaries, tests and observability, not prompts alone.
03
A workshop is only as valuable as what the team still practices the following Monday.
04
Small, composable abstractions beat large, clever frameworks over the lifetime of a product.
05
Independent practice is a commitment to writing code, not just talking about it.
05

Based in Frankfurt. Two decades in functional programming, distributed systems and developer tooling. Recognized F# Expert. Regular speaker at BobKonf and other engineering conferences. Publishes long-form technical videos and essays on programming, AI and craft.

I work solo, in small teams, or alongside existing engineering groups. Engagements run from a single workshop day to multi-month advisory retainers. Clients include product companies, research groups and internal platform teams.

  • F# Expert (community-recognized)
  • Speaker, BobKonf
  • Published on YouTube
  • NuGet packages & OSS libraries
  • Cumin & Potato GmbH — co-founder
06

Short brief, rough deadline, and — if possible — one sentence on what “done” looks like. I reply within a few working days.