From a dashboard for one practice to the data engine behind an entire PMS.
How Smile Data turned a single-practice dashboard into a multi-tenant analytics platform
The reporting, the data, even some of the KPIs already existed. What Smile Data needed was a foundation solid enough to build a business on. We rebuilt it from the ground up, and what came out the other side was good enough that some of the leading practices on its practice management system (PMS) signed on as users before the product even launched.
Smile Data started as one owner's attempt to make sense of their own dental practice's raw PMS data. We stayed on as that fix turned into a market opportunity: a full analytics product built for every practice running on that same PMS.
Why Dynamic Data?
Multidisciplinary team that came together to offer a comprehensive service
A rare combination of
technical abilities and
interpersonal skills
Fully customized service,
adaptable to the
client's needs
About Smile Data
Industry: dental and healthcare software
Services: data warehouse architecture, ingestion pipeline rebuild, dbt modeling, embedded analytics
Smile Data is dental analytics software built on top of a leading practice management system (PMS). It started as internal reporting for its founder's own practices and has grown into a standalone product distributed to dental practices well beyond that founder's own.
The process
The goal
Turn raw PMS data into the KPIs that run a dental practice, in a format every role can use.
What started as a single-practice dashboard quickly became something bigger. Building KPI logic that covered everything from day-to-day operations to C-level reporting was only half the problem. Selling the product to other practices meant re-architecting around a multi-tenant data model with two levels built in: each practice's data fully isolated from every other practice's, and, within a practice, its individual clinics rolled up cleanly into one organization-level view instead of duplicated or siloed numbers.
The stack
Fivetran · BigQuery · dbt Cloud · Cube · Looker Studio
We built the data layer behind the product across five connected pieces, each with a clear job:
Fivetran pulls raw data directly from every practice's PMS, automatically.
BigQuery stores all of it in one central warehouse instead of dozens of disconnected databases.
dbt Cloud turns that raw data into roughly 200 KPIs, defined jointly with the client to actually run the business, from day-to-day operations up to C-level reporting. Each KPI is guarded by a matching automated test and structured to enforce isolation at two levels: between practices, so competing practices share the same warehouse without ever seeing each other's numbers, and within a practice, so its multiple clinics roll up into one clean organization-level view.
Cube serves that isolated, KPI-ready data as a governed semantic layer to the dashboards, so every query respects the same tenant boundaries by construction.
Looker Studio builds the dashboards themselves, which Smile Data embeds directly into its own product.
That architecture is what makes it possible for Smile Data to run as a real multi-tenant, white-label product, rather than a dashboard rebuilt from scratch for every new customer.
The results
Smile Data secured the leading practices running on that same PMS as distributors, signed during beta, before the product's public launch.
Competing practices now run securely on the same platform, with each practice's data fully isolated from every other's.
What started as an internal reporting dashboard became one profitable product, sold across the market.
Delivered on schedule even as the product strategy kept shifting along the way.
Ash Toub
“What we ended up creating is the best product in a very competitive market, to the point that, prior to launch, during beta, we signed contracts with the biggest potential distributors of the product, including some of the leading practices on that same PMS. That's rare.”