Jan Schäferjohann @JanFS

Building customer communication management solutions @kwsoft. Sharing insights on tech, innovation, and smarter systems. https://t.co/g0HGgA2786 Bremen, Germany Joined August 2008

Tweets

612
Followers

149
Following

226
Likes

237

Jan Schäferjohann @JanFS

19 hours ago

Shipping inside big orgs is weird: the work takes ~8 hours, the calendar takes ~6 months. Why? Ever‑increasing specialization + outsourced seams = no one owns the path to prod. My playbook: • Pick a mode: prepare deeply or communicate fast • Bilateral connections > big…

0 0 0 9 0

Jan Schäferjohann @JanFS

22 hours ago

The amount of dev setups I see with one 32 inch instead of 2x27 is confusing to me. Why?

0 0 0 15 0

Jan Schäferjohann @JanFS

2 days ago

I really do need individual mcp support for @ChatGPTapp Teams plans. My workflow is different than others and I need to choose my tools individually. It is the one thing I am really preferring in Claude Desktop. Please let me use gpt-5 fully and connect to my tools.

0 0 0 31 0

Jan Schäferjohann @JanFS

2 days ago

Our CCM application is deeply embedded in existing IT and business workflows. The resulting system lives and changes with your business. Customer communication is much more than just the output produced at the end of the chain. It is the representation of how an enterprise has…

0 0 0 18 0

Jan Schäferjohann @JanFS

3 days ago

The hardest part of CCM isn’t the tech. It’s navigating architecture questions, brand typography, marketing asks, and LOB process rethinks—often at once. It is always interdisciplinary. Cross‑discipline leadership wins; experience is the multiplier.

0 0 0 25 0

Jan Schäferjohann @JanFS

a week ago

How much ARR does this company make? 1 Product 5 Users on a monthly plan at 1 USD each 2 Users on a yearly plan at 5 USD each Professional services sold this year for another 5 USD Whats the ARR?

0 0 1 12 0

Jan Schäferjohann @JanFS

2 weeks ago

GPT-5 is actually awesome, no way around it.

0 0 0 20 0

Jan Schäferjohann @JanFS

2 weeks ago

Written up some of my lessons in using AI tools for generating code over the years in my recent blogpost. Here are the ones that stuck: • Start tiny — ship a script, not a system • Tight loops > perfect plans: hot reload + frequent deploys • Treat the LLM like a collaborator:…

0 0 0 23 0

Jan Schäferjohann @JanFS

3 weeks ago

Is there a way to get a detailled breakdown of potential deployment issues from @vercel? Going 1 by 1 seems a bit weird

0 0 0 23 0

Jan Schäferjohann @JanFS

3 weeks ago

Using @kirodotdev first time today: 1) For clearly defined tasks the workflow approach makes it even easier to get things done quickly. 2) @claudeai Sonnet 4 is awesome still.

0 0 0 22 0

Jan Schäferjohann @JanFS

4 weeks ago

Hey GPT5 in Cursor. Can you add logging to my vercel app? Find my requirements in logging.md, also here is vercel documentation: COPYPASTE 3 Minutes later: Yall can hate vibing all you want, I love it.

0 0 0 71 0

Download Image

Jan Schäferjohann @JanFS

4 weeks ago

Really dont like this take. Would we not want models that can be used expertly?

Casey Flint @FlintCasey

4 weeks ago

Really dont like this take. Would we not want models that can be used expertly?

4 6 175 15K 4

1 0 0 123 0

n8n.io @n8n_io

4 weeks ago

Hamburg 👋 We’re proud to sponsor AI Tinkerers #3, with hands-on demos & builder networking happening this Thursday. AI Tinkerers is a meetup for practitioners actively building and working with foundation models (LLMs & generative AI). Got something to show? Grab a 5-min demo…

3 6 40 5K 5

Download Image

Jan Schäferjohann @JanFS

4 weeks ago

The ability of LLMs to live with my typos is astonishing. At the same time, it already introduces a lot more typos into my other communication as well...