Modern reporting engine and AI data infrastructure for a multi-channel retailer on top of legacy SQL systems

Saint Bernard operates seven physical retail stores and one online store. The physical stores feed into an internal legacy system backed by Microsoft SQL Server. The online store runs on Shopify. Two separate data worlds, no unified view.

Their existing MSSQL database could not be replaced. The physical store operations depended on it. But it was old, the table structure was not clean, and building any modern reporting or automation layer on top of it was not viable. The reporting tools they had were equally outdated. Brand representatives and store managers had no clean way to see the data relevant to them. Leadership had no consolidated view across all channels.

Solverhood built a self-hosted custom ETL pipeline that continuously syncs data from the MSSQL system into ClickHouse. Approximately 20 million rows are processed daily. During the transfer, the data is transformed: the scattered, inconsistently structured legacy tables are reshaped into a clean, well-organized schema inside ClickHouse.

Shopify data from the online store feeds into the same ClickHouse environment, creating a single unified data layer that covers all seven physical stores and the ecommerce operation together.

On top of ClickHouse, Solverhood built a modern reporting frontend. Store managers and brand representatives can log in and see exactly the reports relevant to them. Permissions are scoped by store or brand. Leadership sees everything. The final layer is an AI chat interface built using ClickHouse MCP, allowing the team to query their data directly in natural language.

Saint Bernard now has a unified, real-time view across all eight channels. They own the full ETL codebase and the ClickHouse infrastructure outright, with no dependency on third-party ETL vendors. Against tools like Airbyte, this saves them approximately $25,000 per month.

The legacy MSSQL system remains untouched. The ETL layer abstracts its complexity entirely. Brand representatives and store managers have access to the reports relevant to them without needing IT involvement. And with the AI chat interface, the team can ask questions of their data directly, turning a reporting system into an operational intelligence tool.