{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/4c05501e9fa14edd97f10663f32faa7d\" frameborder=\"0\" width=\"1670\" height=\"1252\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":1252,"width":1670,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":1252,"thumbnail_width":1670,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/4c05501e9fa14edd97f10663f32faa7d-4e2f5528f9e4e04e.gif","duration":593.835,"title":"Build a Snowflake Data Warehouse in Minutes","description":"This Loom demonstrates a no-code data warehouse builder that automatically turns uploaded data into a modeled star schema using Autopilot and LLMs. The presenter uploads CSVs or Postgres-derived data, loads it into Snowflake, then runs profiling to classify fields as dimensions or facts and infer relationships like key joins (for example date). After packaging metadata into LLM calls (noting DeepSeek and Flash and that it chose Flash in one run), the tool generates dbt files, builds the warehouse, and produces lineage, star schema views, and a set of reports and a dashboard in under 10 minutes. An optional improve loop repeatedly reviews the ETL code and can be run about five times for better results, potentially taking around half an hour."}