WeeklyTalk #165
WeeklyTalk AMA | Your questions, Our answers
For this episode, we’ve turned the tables: Instead of setting the topics ourselves, we asked our customers—what do you want to know?
For this episode, we’ve turned the tables: Instead of setting the topics ourselves, we asked our customers—what do you want to know? The result is a series of questions about cloud vs. on-premise, Kubernetes in production, AI integration, and the unique aspects of pharmaceutical IT. Our experts provide short, honest answers—no slides, no fluff.
00:01:19 - What distinguishes professional software development from 'it just works'?
00:01:47 - How much time should you invest in testing and when is it too much?
00:02:17 - What is your take on testing for in-house developments?
00:03:57 - Clean Code: Ideal or reality in everyday project work?
00:05:37 - Documentation: How much is enough?
00:06:01 - Kubernetes in one sentence: What is it exactly?
00:06:28 - When is Kubernetes worth it and when is it overkill?
00:07:23 - Kubernetes: Customer requirement or internal decision for the client?
00:07:58 - Kubernetes in production: What is the most difficult part?
00:08:29 - ChatGPT in daily work: Helpful or a dangerous generator of half-knowledge?
00:09:02 - AI and data privacy: How do they fit together?
00:09:50 - Do our customers have the topic of data privacy in the cloud on their radar?
00:10:38 - RAG-Pipelines: Which embedding models and vector databases do you use?
00:11:29 - Will AI replace developers?
00:12:05 - How do you systematically evaluate the quality of LLM outputs?
00:13:27 - Which technology is currently overhyped?
00:14:28 - Power Platform in a GXP environment: What is possible?
00:15:46 - Most common mistake made by IT professionals without pharma experience?
00:16:32 - Working agile despite Change Control: How do you achieve that?
00:18:35 - What was the trickiest technical decision in the last year?
00:19:45 - What advice would you give your younger developer self?
April 09, 2026