Introducing Tinybird Code: The AI ClickHouse® expert for your projects. Learn more.Learn more
Back

Tinybird Customer Story

How Sync bills for real-time AI model usage with Tinybird

Learn how this growing AI video lip-sync startup built customer-facing usage dashboards and internal analytics to power their transition to usage-based pricing, going from proof of concept to production in less than a week.

Sync needed real-time visibility for usage-based pricing

Sync is an AI research company building models to lip-sync humans in video, backed by Y Combinator and Google Ventures. With customers ranging from individual creators to enterprise media companies, they serve their technology through both an API and a studio playground interface.

As the company grew, they faced a critical business transformation: moving from a simple credit-based system to usage-based pricing that would scale with their diverse customers.

We were transitioning from a credit-based system, where you top up your account, to usage-based pricing, based on the number of generate frames with our models. We needed to start with a system to let our customers query that data easily.

Tair AsimHead of Product at Sync

This wasn't just about internal metrics. Sync wanted to give customers transparency into their usage - particularly important when dealing with video generation where users track frames generated, costs per generation, and usage patterns across different time periods.

<2
days toproof of concept
<1
week toproduction
3
personteam

Performance requirements ruled out traditional databases

As a small but fast-moving startup, Sync initially considered using familiar technology. However, they quickly realized their usage data requirements didn't fit the traditional database model.

It's a performance thing, and the relational database wouldn't make sense. It's an append log of usage data, and the amount of data is massive.

Tair AsimHead of Product at Sync

The team had experience with ClickHouse from previous companies and knew it was the right technology for their analytics workload. But as a small team of three, they couldn't afford the complexity of self-hosting ClickHouse - or even building additional infrastructure to integrate other hosted options.

We knew that the data would grow. We have experience working with ClickHouse, and we knew that we wanted to use ClickHouse, but we didn't want to host it.

Tair AsimHead of Product at Sync

Tinybird delivered ClickHouse® power with startup simplicity

The path to Tinybird came through a trusted recommendation from another successful founder.

I heard about Tinybird from a founder of a really successful startup. I just saw him talking about using it, how it works really well on massive amounts of data, and it was kind of a no-brainer. Tinybird makes it so easy to query the OLAP database from our application.

Tair AsimHead of Product at Sync

Several factors made Tinybird the perfect fit for Sync's TypeScript-heavy stack:

  • Community-supported TypeScript integration: The team uses open-source packages that provide Zod validation over Tinybird pipes, making integration seamless with their existing codebase.
  • Git-based workflow: Everything could be version controlled, avoiding "ClickOps" and maintaining their infrastructure-as-code approach.
  • Intuitive pipe concept: The abstraction of pipes created clean separation between data and application layers, allowing the team to iterate quickly.

From concept to production in record time

With their small product team - just three people including Tair - speed of implementation was crucial. Tinybird delivered beyond expectations.

I think the main thing for us was just getting this out of the door as fast as possible. Because we are a small startup, we are moving fast to gain that advantage. Tinybird was very easy to just set it up and integrate with our app.

Tair AsimHead of Product at Sync

The implementation timeline was remarkably fast:

  • Proof of concept: Less than 2 days
  • Production deployment: Less than 1 week
  • Team size: 3 people (with no prior Tinybird experience on the team)

Sync built two major use cases

Customer-facing usage dashboards

Sync built comprehensive billing dashboards where customers can:

  • View usage by generation (frames generated, costs)
  • Filter by different time periods
  • Track usage patterns month-over-month
  • Resolve billing disputes with detailed per-generation data

Internal billing analytics

In an innovative second use case, Sync pipes Stripe webhook events into Tinybird, creating a centralized billing data warehouse. This powers their Metabase dashboards for daily business intelligence including:

  • Subscription metrics
  • Revenue tracking
  • Financial KPIs for investor updates

Our CEO still uses that data every month to send investor updates. So we have these dashboards in Metabase, which are all powered by data stored in Tinybird.

Tair AsimHead of Product at Sync

Building for the future with the right foundation

For Sync, Tinybird has become more than just a usage tracking solution - it's a key piece of their data infrastructure that scales with their ambitions. The platform team is already planning additional use cases, confident in the foundation they've built.

When asked about advice for teams facing similar challenges, Tair emphasizes the importance of choosing the right tool from the start:

If you need good tooling to get up and running with ClickHouse, I think Tinybird is the right choice. You can experiment with your data pipelines before actually integrating them, so it's easy to validate really quickly.

Tair AsimHead of Product at Sync

Fast-moving startups building analytics features and products, like Sync, must focus on shipping features, not managing infrastructure. With Tinybird, you can get both the performance of ClickHouse and the development velocity that you need to win - even with a three-person team and one week's time.

Want to learn more about Tinybird?
We'll show you how to build your first real-time analytics API in just a few minutes.

Discover the power of real‑time analytics

Tinybird wordmark