Koyeb is a serverless startup that ingests, processes and stores data with multiple cloud providers

Meet Koyeb, a new French startup founded by Yann Leger, Edouard Bonlieu and Bastien Chatelard who have previously worked at Scaleway for many years. Koyeb is a serverless startup that helps you manipulate data in different ways without worrying about your server infrastructure.

Competition has become incredibly fierce between cloud service providers, and Koyeb wants to take advantage of that. You can integrate Koyeb with multiple cloud service providers and let Koyeb do the heavy lifting.

For instance, you may store a ton of videos on an object storage bucket managed by DigitalOcean. Let's say you want to re-encode those videos to optimize them for a new device. Koyeb can import data from this bucket, re-encode those videos and upload the new files to your bucket.

But Koyeb goes one step-further by letting you mix and match services and APIs. As cloud platforms become smarter, they provide services that go beyond running servers and storing data for you.

For instance, Google Cloud's speech-to-text API is arguably better than Amazon Transcribe. Instead of having to manually set up a multi-cloud workflow, Koyeb can take video files from an AWS S3 bucket, transcribes the audio from those video files on Google Cloud and save the result on the AWS S3 bucket.

There are many use cases for Koyeb. It ranges from copying data from an S3-compatible object storage provider to another every day for redundancy to triggering data processing with API calls. Everything scales automatically and once a workflow is done, you no longer get billed for runtime.

There are already dozens of integrations with data sources (as input and output) and ready-to-use processing APIs. Everything can be configured in the web interface with multiple processing steps, using a command-line interface or the Koyeb API.

The company is just coming out of stealth and is already working on more product updates. For instance, you'll be able to use Docker containers and custom functions in the future, which should enable a lot more workflows. But it's a promising start.