Didn't know the project, but going to test soon! Very cool. One question would be if there is already implementation of (or plans to add) any LLM to automatically analyze errors and report them back?
@andrew_correa Are you referring to something like an AI SRE? or more along the lines of anomaly detection?
Report
Maker
@andrew_correa Great question! These capabilities are some of the most exciting on our roadmap. Have you started to integrate LLMs with your observability strategy today?
@jacob_swiss unfortunately im still doing it like the ancestors and manually checking problems, its a particular with cloudflare workers tbh, grafana is easy to use and search stuff, convex with mcp is also easy to plug into Claude Code, sentry I havent tested tho.
Report
Maker
@andrew_correa Manual checks are a soul crusher especially with Cloudflare Workers. If you like the Grafana vibe but want something way lighter check out O2.
Since you're already using Claude Code and MCP are you trying to pipe logs into your AI workflow for auto-debugging or just looking for a decent dashboard?
Report
Maker
@andrew_correa we already support llm observability , llm evaluations will be in preview soon.
Report
This looks like a solid alternative to the big players! I’ve been looking for an observability tool that doesn't feel like a second job to manage. The 2-minute setup claim is bold—I’m going to spin up a cloud instance today and see how it handles my logs. Excited to see where the AI SRE agent goes!
@new_user___07720267fad0ea729d0e1af You can try to turn OpenObserve on a cloud instance. However to start, just try following this doc on your laptop, and in under 30 seconds, you will have OpenObserve up and running with complete GUI, logs, metrics, traces, dashboards, and more - https://openobserve.ai/docs/getting-started/
You can use self-hosted or cloud - Either way, you will be up and running in 30 seconds.
Hey everyone. We built OpenObserve to fix the headaches of complex observability, but the community is what actually shaped it. The GitHub stars, the feedback, the bug reports, the honest conversations.. we owe our continued success to you.
I'll be hanging out here - happy to answer any questions!
Report
Maker
Hi everyone! 🙋 It's been a wild journey moving from our first lines of Rust code to this launch. OpenObserve was really shaped by our early open-source contributors who pushed us to make it API-compatible with things like Prometheus and OpenTelemetry.
We’re just getting started with our AI-native features, and I’d love to hear from this community: how can we make your on-call rotations or debugging sessions less painful? Drop your questions below!
Report
Maker
Expanding on our AI-Powered Insights: We’re big believers that AI should be a co-pilot, not a black box. While our automated SRE agent handles the heavy lifting of the analysis, we make sure to present every correlated signal. It’s all about keeping the human in the loop—giving SREs the full context they need to make fast, informed decisions with total confidence.
Report
The NLP-to-SQL angle is the part that stands out. Most log tools let you query, but translating an engineer's question into the right syntax is still where triage time gets burned. Natural-language querying on top of existing log streams feels like a practical win, and the DAG view for LLM trace chains is a smart addition.
Monostate AItraining
Didn't know the project, but going to test soon! Very cool. One question would be if there is already implementation of (or plans to add) any LLM to automatically analyze errors and report them back?
OpenObserve
@andrew_correa how are you? Great question. The answer is yes there is. Here are some more details:
https://openobserve.ai/docs/integration/llm-applications/#configuration
Monostate AItraining
@shohams I think this link is for LLM observability, I meant to ask about using LLM to observe logs itself tho
OpenObserve
@andrew_correa Are you referring to something like an AI SRE? or more along the lines of anomaly detection?
@andrew_correa Great question! These capabilities are some of the most exciting on our roadmap. Have you started to integrate LLMs with your observability strategy today?
Monostate AItraining
@jacob_swiss unfortunately im still doing it like the ancestors and manually checking problems, its a particular with cloudflare workers tbh, grafana is easy to use and search stuff, convex with mcp is also easy to plug into Claude Code, sentry I havent tested tho.
@andrew_correa Manual checks are a soul crusher especially with Cloudflare Workers. If you like the Grafana vibe but want something way lighter check out O2.
Since you're already using Claude Code and MCP are you trying to pipe logs into your AI workflow for auto-debugging or just looking for a decent dashboard?
@andrew_correa we already support llm observability , llm evaluations will be in preview soon.
This looks like a solid alternative to the big players! I’ve been looking for an observability tool that doesn't feel like a second job to manage. The 2-minute setup claim is bold—I’m going to spin up a cloud instance today and see how it handles my logs. Excited to see where the AI SRE agent goes!
@new_user___07720267fad0ea729d0e1af thank you for your feedback, not only setup is easy , our upgrades are seamless.
@new_user___07720267fad0ea729d0e1af You can try to turn OpenObserve on a cloud instance. However to start, just try following this doc on your laptop, and in under 30 seconds, you will have OpenObserve up and running with complete GUI, logs, metrics, traces, dashboards, and more - https://openobserve.ai/docs/getting-started/
You can use self-hosted or cloud - Either way, you will be up and running in 30 seconds.
@new_user___07720267fad0ea729d0e1af Appreciate it!
Let us know how it goes, especially the setup and logs.
Hey everyone. We built OpenObserve to fix the headaches of complex observability, but the community is what actually shaped it. The GitHub stars, the feedback, the bug reports, the honest conversations.. we owe our continued success to you.
I'll be hanging out here - happy to answer any questions!
Hi everyone! 🙋 It's been a wild journey moving from our first lines of Rust code to this launch. OpenObserve was really shaped by our early open-source contributors who pushed us to make it API-compatible with things like Prometheus and OpenTelemetry.
We’re just getting started with our AI-native features, and I’d love to hear from this community: how can we make your on-call rotations or debugging sessions less painful? Drop your questions below!
Expanding on our AI-Powered Insights: We’re big believers that AI should be a co-pilot, not a black box. While our automated SRE agent handles the heavy lifting of the analysis, we make sure to present every correlated signal. It’s all about keeping the human in the loop—giving SREs the full context they need to make fast, informed decisions with total confidence.
The NLP-to-SQL angle is the part that stands out. Most log tools let you query, but translating an engineer's question into the right syntax is still where triage time gets burned. Natural-language querying on top of existing log streams feels like a practical win, and the DAG view for LLM trace chains is a smart addition.
Congratulations
Thank you so much@madalina_barbu
@madalina_barbu Thanks a lot!
@madalina_barbu Thanks a lot!
@madalina_barbu Thanks a lot !