Semantic Layer

Honeydew is a Semantic Layer built on top of Snowflake.<br/>Honeydew makes sure that any business metric, such as MRR, Churn, or Daily Active Users, can be defined once and calculated the same way for every consumer.

How does it work:

1) Customer onboards with Honeydew and sets up their account (find more details [here](https://honeydew.ai/docs/initial-setup)).

2) Customer imports their existing semantics into Honeydew. These can come from a BI tool (e.g. Tableau workbooks, or LookML models), from the ETL (e.g. dbt models, SQL queries or stored procedures code), or just be newly created on top of existing tables in Snowflake.

3) The semantics can be consumed in a number of ways:

- A JDBC proxy, that can support any BI tool, send queries to the Honeydew Semantic Layer. Honeydew will translate these queries to the actual Snowflake queries. Snowflake will execute these queries and the results will be provided back to the BI tool.

- Creating dynamic datasets based on the semantics. Each dataset can be then deployed to Snowflake, as a view or a table, or exported as a a dbt model.

- Querying the semantics directly from within Snowflake, using the table functions and procedures provided with the Honeydew Native App.

For pricing, see https://honeydew.ai/pricing

搜集汇总

背景与挑战

背景概述

Honeydew是构建在Snowflake上的语义层工具，支持用户集中定义业务指标（如MRR、流失率等）并确保计算一致性。它提供JDBC代理、动态数据集生成和Snowflake原生函数三种方式供下游消费，支持从BI工具或ETL系统导入现有语义逻辑。