# What AI agent stack do you actually recommend for enterprise use? — retained record

> Retained reddit record from reddit:AI_Agents.

Canonical: https://highsignal.app/data/records/575046d07780816c

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:AI_Agents)  published 2026-10-13 15:04 ingested 2026-10-14 02:36

[original source ↗](https://www.reddit.com/r/AI_Agents/comments/1x4z2hh/what_ai_agent_stack_do_you_actually_recommend_for/)

## retained text



```
Hi everyone, I'm actually looking into building AI agents for enterprise use cases and trying to figure out what a realistic production - ready stack actually looks like

When thinking about ENTERPRISE the main requirement we keep running into are

 

\* Data privacy, security and action level permissions ( can't just hand an agent broad API keys) 

\* Scalability and relability for production workloads 

\* Observability ( logging, tracing, monitoring and evals)

\* Integration with existing systems and enginering context ( services , owners, environment - we have looked at Port/ internal catalogs for this so agents have live context)

I see framework like langGraph, crewAI, autoGen, and  LlamaIndex mentioned constantly, but it feels like the orchestration framework is only one part of the puzzle 

For anyone running agents in production accross enterprise teams: what does your actual end to end stack look like? 

Which tools have held up, and what did you end up having to build yourselves? 

Any personal experience pros/ cons, or recomendations would be highly appreciated 
```



document fetched 2026-10-14 00:19 · canonical https://reddit.com/r/AI_Agents/comments/1x4z2hh/what_ai_agent_stack_do_you_actually_recommend_for
