I built Interview Tree because synthesizing user interview transcripts by hand is slow and error-prone. Reading through long conversations to find recurring problems, group themes, and pull supporting quotes takes hours. I wanted to see if Claude could do that extraction reliably enough to be useful in a real research workflow.
What it does
Paste any user interview transcript into the editor, hit analyze, and the app returns a structured opportunity tree: a hierarchy of problems, themes nested under each problem, and verbatim quotes from the participant anchored to each theme. The output renders as a collapsible, depth-indented, color-coded tree you can explore at every level.
Every theme is backed by exact quotes — not summaries or paraphrases — so you can see what the participant actually said rather than what the model inferred. There's also a sample transcript preloaded if you want to explore the output immediately without any data of your own.
How it works
