Propel

Accelerate Your AI and Analytics with Serverless ClickHouse®

Freemium

Description

Propel is a comprehensive data platform engineered to accelerate the development of AI and analytics applications. At its core, Propel offers a serverless ClickHouse® backend, renowned for its blazing-fast query performance, which empowers developers to build sophisticated data features without managing complex infrastructure. The platform is designed to simplify the entire data workflow, from ingestion to visualization, enabling teams to deliver insights and data-driven experiences quickly.

It allows users to effortlessly ingest data from diverse sources, including streaming systems like Kafka and data warehouses such as Snowflake or Amazon S3 Parquet files. Developers can then deploy low-latency data APIs, build custom embeddable dashboards using an open-source React UI kit, and define metrics consistently via a semantic layer. Propel also provides robust multi-tenant access controls, ensuring data security and compliance when building analytics into SaaS applications or sharing data with customers.

Key Features

Serverless ClickHouse®: Blazing-fast, fully managed, auto-scaling analytics backend for real-time queries.
Data APIs: Instantly deploy low-latency APIs (Counter, Data Grid, Leaderboard, Metric Report, Time Series, Top Values, SQL) for various data access patterns.
Embeddable UI Components: Open-source React UI Kit with components like TimeSeries, PieChart, and Leaderboard for building custom, embeddable dashboards.
Semantic Layer: Define key business metrics (SUM, AVG, COUNT, etc.) once and query them consistently across different interfaces.
Multi-tenant Access Controls: Securely deliver analytics in SaaS applications with row-level, column-level permissions, and scoped API credentials.
Built-in Ingestion Pipelines: Ingest real-time streaming and batch data from sources like Kafka, Snowflake, S3, Webhooks, and sync from data warehouses/databases.
SQL & GraphQL Interfaces: Query data using standard SQL (via API or PostgreSQL-compatible interface) or GraphQL for flexible data retrieval.
Developer Tools: Includes GraphQL API Explorer, Retool Integration, Admin API, and Terraform Provider for efficient development and management.

Use Cases

Customer-facing analytics
Deploying low-latency Query APIs for applications
Building SaaS dashboards
IoT and operational insights
Product usage metering and reporting
LLM usage analytics and tracking
Self-service custom reporting for customers
Secure data sharing via SQL interface

Frequently Asked Questions

What is data read by queries?

Data read by queries is the volume of data Propel needs to scan to answer a query. The amount depends on the size of your dataset and the complexity of your query. For example, a query for the last 24 hours on a small dataset will read less data than a query for the last five years on a large dataset.

What is data written?

Data written is data that you write into Propel Data Pools in the form of: Data ingested into Data Pools. Data updated or deleted in Data Pools. Data inserted by Materialized views into Data Pools.

Is storage billed per compressed data?

Yes. You can expect significant compression depending on the data.

How can I see the data read by a query?

The easiest way is to sign up for the free tier and build a proof of concept. In the API Playground, you can see how much data is read by each query you run.

Do you have volume discounts?

Capacity commitments give you volume discounts. Contact sales for more information.