trojans1290

Cool. But honestly all you need is mcp, a semantic layer and let your team run their analysis. Then use something like evidence for bi reporting via duck db. Costs almost nothing.

What would you choose for a backend nowadays then?

What was your preferred stack?

How are you managing the. Workflows? Celery? And how do you organize your Django app with so many workflows.

Also doing the same with pydantic ai. Haven’t used open router though. Will check it out.

Is Django still a solid choice in 2026 for new backend projects? Have loved the orm, admin, etc.

So you’d prefer one file with all models? Do you break out into apps for business logic? Keep models in one app/file. Then domain driven design elsewhere?

Can you share more on this? Is it better to keep all models in one app?

Can you elaborate more on this type of role? Stack? Etc.

What are these skills?

Could you share some of the edges you can into? Agreed that seems the hip move now is to go all in on ts without looking at alternatives.

Thanks for the breakdown on the above.

I’m wondering - have you ever tried a batteries included backend framework + react? So maybe ror, Django, laravel, spring boot, etc? At what point is it too much racking on various node packages vs just using a proven backend framework + openapi? Then run wild with react on the frontend.