A data analysis engine that reduces the number of errors made by programming agents
What Exmergo does is “equipping data stacks with AI.” Its main product, Dex, is an open-source analysis engine designed specifically for programming agents. It enables these agents to first understand your data model and semantic structure through Dex before creating metrics or writing analysis logic, thereby reducing the likelihood of them incorrectly interpreting fields or calculating wrong indicators. The developers conducted comparisons on the ADE-bench platform in dbt Labs and found that the error rate decreased by approximately one-quarter as a result of using Dex.
Another product, Viz, acts as a semantic-level visualization tool that allows business users to create visual reports on their own, making it comparable to tools like Tableau and Power BI. It supports various common data sources, including Snowflake, BigQuery, Databricks, ClickHouse, and PostgreSQL.
Dex is open-sourced under the Apache 2.0 license and can be run locally, while Viz is a paid SaaS service. It is ideal for data teams that already have a data warehouse and dbt tools and wish to incorporate AI into their analytical processes.