Drop two files, pick the key column on each side, choose inner / left / right / full outer — get one combined file plus a match report. It's a real hash join running locally; the equivalent SQL is one click away.
Spreadsheets make combining two files miserable — VLOOKUP gymnastics, copy-paste, prayer. This is the database way: declare the keys, pick the join type, done.
CSV, TSV or Excel on each side. The key columns are auto-suggested when both files share a column name.
Pick the matching column on each side (add a second key for compound matches). Inner keeps matches only; left keeps all of A; right all of B; full outer keeps everything.
Choose which B columns to carry over, check the match-rate report, then download CSV/Excel or open the result in the dashboard.
The semantics are exactly SQL's — because that's what people actually need.
Inner, left, right and full outer — including the unmatched-rows behavior each implies, reported honestly in the stats.
Match on two or three columns at once (region + date, id + type) when one column isn't unique enough.
Matched A rows, unmatched on both sides, output count — so you see immediately when a key is wrong.
Every join shows its equivalent DuckDB SQL — copy it into the SQL tool for anything more exotic.