Evidence

@avloss that comes out of the box! We have a library of visualization components. Here’s the docs page for that one, but there’s a full list in the sidebar of the docs site as well: https://docs.evidence.studio/components/calendar_heatmap

@alexcloudstar it’s definitely nice to have everything flowing through version control.

For row-level security in embedded, you pass encrypted user variables through with your request to our Embed API, and those user variables can be used to create row-level security rules in Evidence - so for example, you could pass through a customer_id and have an RLS rule that allows only the data that matches that customer_id

Hope that helps. Happy to answer any questions about it!

@laander that is what’s meant by embedded! There’s really 2 main options people go for with embedded analytics - embedding a BI tool or building it custom. The BI tools are the drag and drop interfaces. Building it custom as it sounds like you’ve done gives you a lot more control but requires an investment to build and maintain those components, interactivity, performance, etc.

For a lot of teams who have the capacity and capability, custom is the right option because you can build exactly what you need. But many of the teams we talk to realize it’s not worth the lift because analytics isn’t their core competency (usually after they try wrangling a JavaScript charting library…). So they reach for Evidence to get everything they need out of the box and still get the flexibility of working in code.

What do you use to build your components?

@jaydev13 you can write SQL directly in your reports, and then reference other queries using jinja like syntax `select * from {{my_other_table}}` and Evidence automatically resolves the final query for you

you can also build "models" as reusable SQL tables that you can use across all your reports

there is no direct "lineage" feature to drill back through to see the tables that made it from an end-user perspective yet. is that something that would be useful to you?

@mykyta_semenov_ thanks! Yes it should work well for building reports on top of e-commerce data. We will be adding specialized connectors for that in the new year (e.g., Shopify, Stripe)