Article

Announcing Airbyte Cloud

Talia Moyal

•

April 5, 2022

•

4 min read

Summarize with ChatGPT

Summarize with Perplexity

Airbyte’s mission of powering data movement applies to every single application that touches data, and the people building, operating or using those applications today. This means, requiring our platform to be easy to use, even by someone who doesn’t know what ELT means.

Why we built it

Airbyte was created as an open-source data integration platform to sync data from data sources like applications, APIs and databases to destinations like data warehouses and data lakes.

After two years of an incredibly involved and excited open-source community, we wanted to understand where we could be better for those less technical, or those with less resources (time and knowledge of setting up data integrations).

These are some of the concerns we set to address:

My teammates want to make data-informed decisions, how can they sync data from the tools they manage?
How can multiple business units manage their own data pipelines?
Is there a single click way for me to set up a data integration?
How can I optimize the cost of data replications?

And finally, not a question rather a statement that captures the sentiment of everything above in a single sentence: I have no idea how to deploy, manage or maintain the infrastructure for a data integration pipeline.

Introducing Airbyte Cloud

Today we introduce Airbyte Cloud – the hosted version of our open-source data integration platform, and the foundation for any person interacting with data to have a seamless experience moving, accessing and analyzing data from the various applications they interact with.

Airbyte Cloud enables saying goodbye to spinning up Docker containers or thinking about scaling infrastructure. We’ll manage everything under the hood for you, so that you can manage growing your business.

Similarly to Airbyte OSS, Airbyte Cloud enables users to move your data in minutes using one of our 100+ pre-built connectors, but with a shorter getting-started time.

Airbyte Cloud lets you and your teams:

Leverage OAuth to connect sources and destinations to bypass finding credentials for every application
Invite unlimited users to a workspace
Create multiple workspaces corresponding to various business units
Optimize cost for high volumes of data transfers
Participate in enterprise support – guaranteed response times and a dedicated team to help along the way

Who should use Airbyte Cloud?

This product is best for anyone who quickly needs to access and integrate data from applications and doesn’t have the time or expertise to build infrastructure to support a data pipeline.

This is for someone who is:

Looking to visualize your sales pipeline by pulling in data from multiple sources like Salesforce and Marketo
Understanding and tying revenue to campaign performance
Creating a full view of what’s impacting P&L
Centralizing access to data from various databases / APIs

A note on cost optimization

Airbyte’s pricing is not novel to the developer industry, but it is novel for the ELT space. And we’re okay with this because our vision is much greater than just ELT / ETL. We’re here to power data movement as an industry, and that requires bringing down the costs in a way that makes sense.

Data movement has historically been prohibitively expensive. Airbyte Cloud follows an infrastructure model, meaning we charge on resources used (or credits). This allows for much more predictable and transparent pricing, meaning you can replicate high volumes of data, such as databases, without worrying about breaking the bank.

Consumption of resources is measured as an Airbyte credit, where the standard price of a credit is $2.50. The number of credits required to run a data sync depends on the data source. There are three data sources:

APIs: 1 hour of syncing costs 1 credit ($2.50)
Databases: 1 hour of syncing costs 10 credits ($25)
Files: 1 hour of syncing costs 4 credits ($10)

We only charge on credits that provide value. This means never charging on failed jobs or job attempts.

Let’s ground this in an example:

Imagine your setup consists of data coming from a Postgres database (source) and a few applications like Stripe or Salesforce (more sources). You want to move that data to Snowflake, your data warehouse (destination). Your Postgres database has about 1,000,000 records which take about 8 minutes to sync. Your API sources have several thousands of records that take about 5-6 minutes to sync. You also know you want to track the change history of your data over time.

In a traditional ETL pricing model, providers charge on the number of active changed rows per month. In our example above, this would cost you around $1,000 and just tracking your change history, which never changes after it’s created, accounts for a third of that bill.

In contrast, with Airbyte’s pricing, this would cost roughly $112.