Commented on Supaboard AI · Oct 19, 2026
Hey DevHunt, Subh here, co-founder of Supaboard. Before this we were the data people at Fortune 500 teams and startup labs. Same job everywhere: write the SQL, build the dashboard, and by the end of the day there is one more request waiting. The tools never did the whole job. One answered questions, one drew dashboards, one scheduled reports, and someone still had to glue it together. So we built one agent that does all of it on your live database. It shows the SQL behind every answer, learns your definitions from corrections instead of a semantic layer, and grades itself against answers you already know. On BI Bench and the other accuracy benchmarks our agents score highest, and in a blind test with 134 people they beat the Claude Code and Codex harnesses. Method is on the site, happy to walk through it here. Three things we think you will like: • Query Bench: SQL and Python editor with tab completion that knows your schema. • No semantic layer, no eval harness to set up. Your corrections become rules, and evals run on every plan. • Postgres, MySQL, SQL Server, MongoDB, BigQuery, Snowflake and any MCP server. Live, no sync. Free plan, no credit card. Viewers are free. On-Premise with your own keys if data cannot leave. DevHunt bonus: code DEVHUNT26 gets you three months of Pro. We are here all week. Tell us what breaks and what you would want it to do next. Subhrajyoti
