Hacker Newsnew | past | comments | ask | show | jobs | submit | pjmalandrino's commentslogin

defenetly a big strugglez.. we are working with researchers on that actually..

I think a descent answer is a cross-factor one.

We have agile teams, that have a velocity measured in story points. This gives a first good idea but it is not the all truth. It is a very relative measurment.

but if you add informations about pure productions : Number of PR's Number of commits in PR's Number of code lines Number of comments on PR's

Then you got something more interesting.

Also, we are looking for a way to add quality indicators to understand if it is just rush, or full vibe coded code that will make project los t in few months ..

But I agree with other comments saying that it is a struggle, and AI coding just make more painful ..


Very impressive series of SLM by IBM here.

I have been using it with their Chunkless RAG concept and it is fitting very well! (for curious https://github.com/scub-france/Docling-Studio)

I convinced that SLM are a real parto of solution for true integrated AI in process...


Why are you building your own DAG system instead of just using LangGraph? You could cut complexity and focus on what actually matters : the claims, evidence tiers, conflict detection.

Also, embedding claims in the Chain of Thought instead of post-processing them might force rigor earlier in the pipeline.

(Assuming the zero-deps constraint isn't a blocker?)


Not same tools, different purpose from my opinion


Wow, very impressive, great job! You mentioned monitoring, I think it might be a very interesting way to see the "ongoing" work of your agents and orchestrate them. Do you have a precise idea on how it's going to happen, or is this already planned?


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: