Build Your Own AI Feedback Loop on Kibana Logs (Detect & Solve Production Issues Automatically)

This hands-on workshop teaches participants how to build an AI-powered feedback loop that monitors application logs in real time, detects anomalies, and automatically assists engineers in diagnosing production issues.

Using Elasticsearch, Kibana, Logstash/Beats, and an open-source LLM (such as Llama 3 or Qwen), participants will:

1. Inject synthetic production issues into logs

2. Create anomaly detection rules

3. Stream logs to an AI inference service

4. Build a pipeline where the LLM reads logs and provides root-cause summaries

5. Configure auto-alerting & routing to Slack/Email

This is an end-to-end practical lab for anyone interested in observability, AIOps, or SRE.

• How to structure logs for machine understanding.

• How to connect Kibana logs → AI model → feedback loop.

• Techniques for prompting, summarization, and root-cause detection.

• Building auto-diagnosis systems using open-source LLMs.

• A working prototype participants can take home and extend.