Data teams already treat queries, models, and transformations as code. Yet embedded analytics is often built in drag-and-drop BI tools, without version control or reliable deployment workflows. That’s a poor fit for one of your most critical, customer-visible data products. Embedded analytics in Evidence introduces a code-first way to build customer reporting that fits seamlessly in your app - and can be owned entirely by the data team.
Hey Product Hunt,
Sean here, co-founder of Evidence 👋
Today we’re launching Embedded Analytics as Code in Evidence Studio.
Embedded analytics is often a company’s most important (and most scrutinized) data product, but it’s still commonly built with drag-and-drop tools, edited outside version control, and deployed without real review or rollback. That approach can work for internal dashboards, but it tends to break down quickly once analytics becomes customer-facing.
We built Embedded Analytics as Code to bring real software workflows to embedded reporting.
With Evidence, embedded reports are defined in SQL + markdown, live in Git, and move through the same review and deployment process as the rest of your app. We also handle the hard production pieces out of the box: row-level security, short-lived embed sessions, custom theming, and performance tuned for customer traffic.
Teams are already using this to deliver analytics to thousands of end users, including professional sports teams and Fortune 500 companies, without turning analytics into a parallel engineering platform they have to maintain themselves.
If you’re embedding analytics today - or thinking about it - we’d love to hear what you’re building. Book a time to chat with us here: https://calendly.com/evidence-st...
Thanks for checking out our launch and for all your support - let us know what you think in the comments!
Sean
Noticed a lot of early teams struggle with support/billing triage —
is this something you’ve already systemized or still mostly manual?
Congrats on the launch! This makes a lot of sense for customer-facing analytics.
This is cool! Are you seeing a trend of every SaaS becoming APi-first? Seems like you're following a great trend.
Wow, Evidence looks amazing! BI as code is exactly what Ive been searching for. Im particularly excited about the version control aspect for embedded analytics. How does Evidence handle complex data lineage across different SQL queries?
hey PH!
we've been working with beta users this year to build a different take on customer facing analytics
existing solutions are either:
drag and drop BI with no release management, and limited customizability
custom JS that the eng team has to own
today we're launching "embedded analytics as code" for the data team
build charts and viz in markdown and SQL
customize the theme to match your app
use the embed API to create single-use, presigned URLs
> Embedded analytics is often built in drag-and-drop BI tools
what's meant by embedded analytics exactly? if it's charts, e.g. in a dashboard, inside my own product that my customers use to consume their own data, then I'd usually built it with components wired up to our own API.
not sure how Evidence fits in here? :)
BI as code hits a nerve. I’m tired of diffing dashboards by screenshot. SQL + markdown feels sane. If embedded reports ship via PRs and not a drag-and-drop maze, that’s a win. Curious about auth/row-level stuff. Feels like the right shape.
Congratulations on your launch! 🚀
So, that `calendar_heatmap` - what does that refer to? Is that some user-defined component, or does it come out of the box with Evidence?
