.svg)
Turn your Kafka streams into analytics APIs in minutes
Tinybird is a serverless analytics backend that Kafka developers use to quickly build fast APIs over streaming data at scale.
.jpg)
How it works
1. Connect
Create a new Kafka consumer in seconds with Tinybird’s native Kafka connector. Just add your server details, select a Topic, and go.
2. Store
Define your schema (or allow Tinybird to guess it for you) and store your events in a managed, columnar Data Source at millions of EPS. Tinybird supports AVRO format and schema registry integration for automatic deserialization.
3. Query
Use Tinybird Pipes to write short, composable SQL nodes to filter, aggregate, and join. Enrich your events to dimensions and materialize on the fly. Tinybird chains and optimizes the nodes for super-fast analytics.
4. Publish
Add query parameters to your SQL with the Tinybird templating language, and instantly publish queries as dynamic, low-latency REST APIs that scale.
2. Store
Define your schema (or allow Tinybird to guess it for you) and store your events in a managed, columnar Data Source at millions of EPS. Tinybird supports AVRO format and schema registry integration for automatic deserialization.
4. Publish
Add query parameters to your SQL with the Tinybird templating language, and instantly publish queries as dynamic, low-latency REST APIs that scale.
“Tinybird is the highly available database that allows you to query quickly and aggressively without having to worry about anything, which is crazy to believe! The abstraction from ClickHouse features, the Data Sources, the Pipes, and how easy it is to deploy an API - one click and instantly available - it’s magic.”
Why use Tinybird?
Native Kafka connector
Storing Kafka events into an OLAP database is simple using Tinybird’s native Kafka connector. With automatic schema guessing and no hassle setup, you’ll be storing events in seconds.
Kafka 🤝 ClickHouse
Kafka deserves a database that can keep up. Tinybird is built on ClickHouse, the analytical database built for ultra-fast ingestion and low-latency aggregation over billions of rows of data.
Completely serverless
Designed to handle traffic spikes at Black Friday-scale, Tinybird can easily ingest millions of messages per minute from Kafka topics and write them to columnar storage for rapid query availability.
As simple as SQL
No Java, no new SDKs or frameworks. SQL is the only skill you need to define metrics and publish low-latency REST APIs from Kafka streams.
REST APIs in a click
It all comes together with the Endpoints. Instantly publish your SQL queries as low-latency, dynamic REST APIs to power your next application or service.
Secure and scalable
Easily generate tokens to control access to your API endpoints, and implement as many access policies as you need, with support for row-level security.
“Tinybird is more than just a managed database and an API service, it’s also an amazing dev tool. The CLI makes it incredibly easy to manage and deploy high scale data pipelines."
Frequently Asked Questions
Can't find the answer you're looking for? Email us at support@tinybird.co or join us on Slack.
How is Tinybird different from Kafka Streams?
Tinybird is a serverless backend service for publishing low-latency analytics APIs from metrics that filter, aggregate, and join stored streaming data. Tinybird stores Kafka data into hosted columnar data tables, and offers SQL as a query language to define metrics that can be instantly published as REST APIs.
Kafka Streams is a client library for building services on top of Kafka, where the input and output data are stored in a Kafka cluster. It provides an abstraction for working with Kafka inside of your code.
So, while both Tinybird and Kafka Streams are both designed for simplifying working with Kafka data, Tinybird is a cloud service focused on building analytics on streaming data as published low-latency APIs, while Kafka Streams is a Java library that you can use to more easily interact with Kafka messages in your code.
How is Tinybird different from ksqldb?
Tinybird is a serverless backend service for creating low-latency analytics APIs on top of streaming data. The purpose of Tinybird is to provide a fast and simple publication layer for analytics on streaming data.
ksqldb is an open-source, distributed streaming processing engine that you can use to analyze and transform data stored in Kafka.
So, while both Tinybird and ksqldb support stream processing concepts, Tinybird is fundamentally a service for publishing analytics APIs on streaming data, whereas ksqldb is designed for stream processing within Kafka.
Is Tinybird a stream processing service?
No, not in the traditional sense. Tinybird consumes data from Kafka topics and other data sources and writes them to a columnar database, then lets you build queries with SQL over the data in those tables. Stream processing services generally focus only on making on-the-fly transformations. Tinybird allows you to create those same on-the-fly transformations using Materialized Views, but with the ultimate purpose of publishing low-latency, high-concurrency APIs.
What kind of message decoding does Tinybird support?
Tinybird supports JSON, AVRO format, and Schema Registry integration for automatic deserialization. Currently Tinybird does not support Protobuf.
Is Tinybird hosted or on-prem?
Tinybird is a serverless backend service hosted on shared or dedicated GCP or AWS infrastructure, depending on your plan. We currently do not offer Tinybird on-premise. To support working with your cloud infrastructure, we can use services such as AWS PrivateLink or Private service access in GCP to create a VPC from your cloud infrastructure to Tinybird.
How much does Tinybird cost?
Tinybird has three pricing tiers: Build, Professional, and Enterprise.
Build plans are free forever, and include 1000 requests per day on your published APIs plus 10 GB of storage.
Professional plans are charged based on data processed by published endpoints and data stored.
Enterprise plans are bespoke to your use case and include premium support and uptime SLAs.
For more information on pricing, check out the Tinybird Pricing page.
Do you offer free trials or PoCs?
We are all for PoCs. If you want to test the limits of Tinybird beyond our Build plan, our customer success team will work with you to get started. Contact Sales to organise one. Challenge us to tackle something you didn’t think was possible :-)
You are also very welcome to start with a Build plan and to keep using it for free as long as you stay within its limits (1000 reqs/day, 10 GB stored). You can upgrade to the Professional or Enterprise plans at any time.
Where does Tinybird store my data?
You can choose between different regions when you set up your workspace. Tinybird is currently available in the U.S. and E.U., but soon you’ll be able to store your data in Asia, South America or Australia.
Can I use Tinybird on-premise?
We do not currently offer Tinybird on-premise. This is in our future roadmap, although we cannot commit to a release date. You can join our Slack to stay informed when we release this.
What type of support does Tinybird offer?
If you are on a free Build plan, the best place to find support is in our Slack community.
We offer support via email in the Professional plan, and also have dedicated customer success engineers and support SLAs for Enterprise plans.
Start productizing your Kafka streams
