Reporting depended on an engineer.
On a US consumer platform, a business question often became an engineering task: filter the data, prepare an export, plot the result. Depending on the question, a report could take hours or days. Decisions waited, and engineering work kept being interrupted.
Build a reporting foundation the team can actually use.
In his lead engineering role, Bhavin built a near-real-time analytics stack with ClickHouse and Apache Superset, including KPI dashboards and scheduled email reports. Superset ran on AWS Fargate.
On top of that foundation, an MCP connector let nontechnical team leads ask questions of live business data in plain English. The language interface only helped because the reporting system could supply the answers.
What changed
Routine reporting moved from hours or days to seconds. Team leads could answer their own questions instead of waiting for a manual export.
That describes the delivered reporting workflows. It is not a promise that every question will return in seconds. Query complexity, source freshness, and metric definitions still matter.
This is the kind of work Developer Labs takes on for clients: agree the metrics, put a dependable data path underneath, and give the business an interface it can use — a dashboard, a scheduled report, or an approved AI tool. You can try a small, fictional version in the lab.
Bhavin wrote a longer engineering account, including query times, on his personal site: From a reporting bottleneck to business answers.