This layer of the stack is all about **presenting data clearly and compellingly** — whether it's to inform decisions, explore trends, or showcase outcomes. Good visualization tools enable both technical and non-technical users to interact with data in meaningful ways. ## What I Value in This Layer - Fast iteration and flexible storytelling (especially for demos or prototypes) - Support for live data, filters, and interactivity - Low friction for non-technical viewers - Role-based sharing or embedding options - Integration with warehouse or other data sources --- ## Tools I Use / Have Used In The Past ### [[Streamlit]] Lightweight, Python-native web app framework for building interactive dashboards, data tools, and internal apps. - Great for prototypes and quick turnarounds - Write everything in Python — no frontend required - Ideal for ML outputs, what-if scenarios, and EDA tools --- ### [[Tableau]] Enterprise-grade BI tool for business dashboards and self-service exploration. - Powerful for chart building, drilldowns, and story dashboards - Works well with structured warehouse data - Preferred for stakeholders used to traditional BI tools --- ### [[Sigma Computing]] Spreadsheet-native BI layer on top of the data warehouse. - Built for analysts comfortable with Excel-like interfaces - Real-time querying on warehouse data (e.g., Snowflake) - Excellent for self-serve dashboards and flexible analysis --- ### [[Metabase]] Open-source BI platform for internal dashboards and metrics tracking. - Easy to set up and collaborate on shared queries - Great for recurring reports and operational monitoring - SQL-native with good controls for non-technical viewers --- ## When This Layer Shines - Demoing results to stakeholders - Enabling self-service insights without writing code - Turning data into decisions, not just tables > **Tip**: Build quick views in [[Streamlit]] for early demos, then move mature metrics into [[Tableau]], [[Sigma]], or [[Metabase]] depending on audience. --- **Part of**: [[Building Data Products]]