We are building Ask Astro in the open as a reference implementation in order to drive a conversation. High-quality orchestration, ingest, monitoring and feedback pipelines are essential to operationalising LLM applications. We want to hear from you.

• How is the Ask Astro demo working for you?
• What are you building with LLMs and Apache Airflow?
• How can RAG-based ingest pipelines be advanced for more scale (larger data, more sources, etc.)?
• Which frameworks, libraries, vector stores are you using and should there be an Airflow provider?
• How can Apache Airflow adapt to make LLM operations simpler, more observable, or scalable.