Lutz Hühnken

@huehnken.de

Fractional engineering leader and software architect. 20+ years in distributed systems & messaging. Helping organizations build scalable, resilient systems and highly effective engineering teams.

(🇩🇪) Ein bisschen Feedback zu meinem EDA-Training. Weil positives Feedback motiviert und weil ich ein bisschen Werbung machen will 🙂 Nächster Termin: 26./27.11. in Hamburg. Thread 🧵👇

New talk: the Kafka ecosystem in 2026. Drop-in alternatives speaking the Kafka protocol (Redpanda, WarpStream, AutoMQ, Bufstream), what diskless architectures actually change, smarter clients, proxies and gateways, plus KIP-932 and KIP-1150. Kafka is a protocol now, not just a broker. #kafka #eda

Banner ad promoting a talk about Kafka at Java Forum Nord 2026 in Hannover.

Part of my onboarding at Typesafe (a.k.a. Lightbend, a.k.a. Akka) was a training with Martin Thompson. I loved it! It's great to see that his thinking in terms of "Mechanical Sympathy" is still being rediscovered and popularized. 👉 martinfowler.com/articles/mec... #mechanicalsympathy #programming

Principles of Mechanical Sympathy

Principles of mechanical sympathy for creating performant software: Memory access patterns, false sharing, the single-writer principle, and natural batching.

martinfowler.com

🐦️ 𝗕𝗶𝘀 𝗱𝗶𝗲𝘀𝗲𝗻 𝗗𝗼𝗻𝗻𝗲𝗿𝘀𝘁𝗮𝗴 𝗧𝗶𝗰𝗸𝗲𝘁 𝘀𝗶𝗰𝗵𝗲𝗿𝗻 & 𝟭𝟬𝟬 € 𝘀𝗽𝗮𝗿𝗲𝗻!🥳 • 16.–17. Juni 2026 im Holiday Inn München-Unterhaching • Themen u.a.: #DDD #API #KI #Cloud #EmbeddedSystems 𝗪𝗮𝗿𝘂𝗺 𝘁𝗲𝗶𝗹𝗻𝗲𝗵𝗺𝗲𝗻? ✔ 23 #iSAQB Expert:innen ✔ 23 Sessions auf Deutsch ✔ Networking & Get-together Jetzt anmelden 👉️ t1p.de/4s17l

Speaker des Software Architecture Forum 2026Speaker des Software Architecture Forum 2026

New version 0.19 of Ollama leverages Apple's MLX framework, ~doubles the tokens/sec for optimized models. Tried it on an M4 Max 👍. It's no Opus, but a nice local alternative that you can play around with without tapping into your token budget. 👉 ollama.com/blog/mlx #claude #ollama #ai #localai

Ollama is now powered by MLX on Apple Silicon in preview · Ollama Blog

Today, we're previewing the fastest way to run Ollama on Apple silicon, powered by MLX, Apple's machine learning framework.

ollama.com

Event-driven architecture is everywhere, but getting it right is still hard. In the next months I'll be speaking about EDA patterns and practices at 3 European conferences: JAX Mainz (May), DDD Europe in Antwerp (June), and a full-day masterclass at WeAreDevelopers in Berlin (July). Let's meet!