Creator of the AI Unified Process. Java Champion, Vaadin Champion, Oracle ACE Pro, Author, Speaker, Programming Architect, and Lecturer.

Erlach, Schweiz
Surprise! My book arrived two weeks before the official release date. 🎉 "Spec-Driven Development: From Specs to Code with AI Agents" (Apress Pocket Guides) is now available, both as softcover and eBook. It is a compact guide on how to make specs the contract between you and your AI agents, based on the AI Unified Process (AIUP). Requirements, use cases, entity models, and a complete case study, all in about 150 pages. I hope it helps you bring more structure into AI-assisted development. 👉 link.springer.com/book/10.10… #SpecDrivenDevelopment #AIUnifiedProcess #GenerativeAI #KeepITSimple
8
32
227
13,523
I’m excited to be a part of SOFTCON 2026! Join me at the Philippines’ premier software conference. softconph.com/
7
562
After many years as a team lead, I know when control matters. That helps me a lot when working with AI. But trust works differently: when someone on my team delivered good work ten times, I could rely on them the eleventh time. With AI, I can't. So I trust specs and tests, not the tool.
3
5
34
1,252
Simon Martinelli retweeted
Says @simas_ch: "“Write the use case before you prompt: write the actors, the main scenario, the alternative flows & the business rules in a simple Markdown file. Give that file to your AI agent and let it implement the feature and the tests from it." 🔗 ow.ly/4rR550ZR08w
1
3
321
In the 90s, I programmed COBOL on the IBM mainframe. Later came client/server. Back then, I had all the roles. I talked to the customers, did the architecture and design, wrote the code, and made sure the application ran in production. Then we specialized. Business analyst, architect, developer, tester, operations. Every handover lost knowledge. With AI coding agents, we are back. One person has time for the whole thing again. But two things are different: 1. The specification is now input for the AI, not just documentation. Good spec, good code. Bad spec, a lot of bad code, fast. 2. Systems are much more complex. The most important skill is judging if the AI result is correct. The new role? An architect who gets their hands dirty again. The tools are new. The principle is old. Keep IT simple. Read the full post: martinelli.ch/back-to-the-90… #SpecDrivenDevelopment #AIUnifiedProcess #SoftwareArchitecture #AI
2
7
53
2,346
Simon Martinelli retweeted
Vaadin 25.3 improves the AI dev loop 🤖 🧠 MCP, skills & agent tools give coding agents up-to-date Vaadin context ⚡ Browserless tests let agents self-verify changes in milliseconds 🎨 Dev Loop makes UI tweaks visible instantly vaadin.com/docs/latest/build…
Made with AI
4
5
11
688
Simon Martinelli retweeted
I look at it differently I think most people I knew got into the industry because they really liked coding. Now that AI is in the process of affecting every part of that industry its not the ego that is being shattered but process of the craft we got paid to do and loved is being shattered. But of course I think there is many things to be excited about and I do think that even though the old way is going to fade out over the years for the majority of programmers, there is still so much amazing things you can do. Not a doomer
Best part of AI is shattering dev's egos, the worst part of this industry. You are not special. Your code actually sucked like everybody else's. You were never the smartest person. The things my sales guy BIL and designer brother are doing is beyond anything I have done.
136
63
1,906
215,275
Simon Martinelli retweeted
Opus 5.5 performs at the level of Fable 5.1. It's ~30% faster and ~40% cheaper than Opus 5 per task. In Claude Code: - 5-hour session limits increase 20% today - Opus 5.5 is priced lower, so it goes 25% further within limits - Pro, Max, and Team users get a reset to use anytime
Introducing Claude Opus 5.5, the first model in our new Claude 5.5 family. It performs at the level of Claude Fable 5.1 for most tasks, and costs 40% less to run than Opus 5.
465
1,204
16,293
1,516,625
Simon Martinelli retweeted
🤖 AI Coding war der Anfang. Jetzt kommen die Agenten. @AdamBien , @simas_ch & @mobileLarson diskutieren auf der W-JAX den Weg vom individuellen AI Coding zum skalierbaren Agentic Engineering. 📍 München | 📅 04.11. | ⌚ 09:00 Uhr 🔗 f.mtr.cool/x7zrp8kpoh #jaxcon
2
2
365
Simon Martinelli retweeted
Now kids can stop complaining about it...
We're adding support for AGENTS.md to Claude Code. Starting today in version 2.1.277, if there is no CLAUDE.md in a folder, Claude will check for and use AGENTS.md. You can toggle this behavior in /config.
1
2
693
I think this is the first time in history that software developers are concerned about (token) costs. What about their salaries, though? Do they think they’re working for free?
5
1
16
1,886
The credit card company resolved the issue with the incorrect charge from @bookingcom , and I received my money back. An answer to my support inquiry? Nothing! I no longer trust these large US companies and will be very careful when using my credit card with them
3
3
14
1,365
Harness Engineering is not about writing bigger CLAUDE.md files. It is about designing the environment in which AI agents work. In my latest article, I explain why a minimal CLAUDE.md and a good Software Architecture Document belong together. The idea is simple: • CLAUDE.md is the map, not the knowledge base • The architecture document contains the actual architectural rules • Skills guide the agent before it starts coding • Tests and ArchUnit act as sensors that prevent violations from getting through The result is a shift from telling the agent what to do to designing a system in which the agent is guided and constrained by architecture. For me, this is where Harness Engineering and Spec-Driven Development really connect. Read the full article: Harness Engineering: Why a Minimal CLAUDE.md and a Good Architecture Document Belong Together martinelli.ch/harness-engine… #AI #HarnessEngineering #ClaudeCode #SoftwareArchitecture #SpecDrivenDevelopment #AIUnifiedProcess
7
13
59
3,152
Simon Martinelli retweeted
Simon Martinelli has spent the last two years applying AI to large #Java systems. At JRush Episode 8, he’ll share lessons from real modernization work: the approaches that helped, the ones that had to be dropped, and the places where AI still needs a careful human hand. 🏆 @simas_ch is a Java Champion and creator of the AI Unified Process. Don’t miss the live session: jrush.bell-sw.com/episode8
2
7
343
Simon Martinelli retweeted
#Java27 has arrived! 🥳☕ Get all the details about the latest release from our Java team. social.ora.cl/6015BGPgHt
1
50
207
5,929
Simon Martinelli retweeted
Truman said it best
258
6,025
21,788
275,687
Whenever I come across the website of a tool or service, I always want to know who is behind it. If that isn't clear, I won't buy anything there. However, many websites nowadays tend to be rather anonymous or vague about their origins. How do you handle that?
4
2
10
1,124
I really like macOS 27 so far! Especially that I no longer need Bartender or ICE for the menubar:
6
17
3,239
Simon Martinelli retweeted
⭐️ One of the most favourited Devoxx Belgium 2026 Deep Dives! @simas_ch dives into Spec-Driven Development: moving beyond prompts to specs that guide AI coding agents from requirements to tested code 🚀 m.devoxx.com/events/dvbe26/t…
2
1
9
837
Simon Martinelli retweeted
What an incredible presentation by Oracle's Ron Pressler on memory management in Java piped.video/watch?v=xr73mR7i…
3
19
145
10,475
Simon Martinelli retweeted
Vaadin Create 2026 — Barcelona, 27–28 Oct. 20+ speakers, 25+ sessions on AI agents in real Java projects, database-to-UI in one language, Spring & Jakarta EE, and legacy modernization. Agenda + 30% OFF tickets: vaadin.com/vaadin-create-202… #Java
5
15
700