Analyse customer interviews, from transcript to decision.

Tallybird interviewed ten customers about month-end reporting. See what they said, read each interview, write up what to build next, and ask about it all from your own AI.

Start with Solo · £9.99

The Brief and its themes

The answer first, then every theme counted by people, each one opening to what people meant.

  • Every theme, counted by people8 of 10 at the top
  • Each theme split into what people meant3 ideas in one
  • What people asked for, in their words3 asks
  • A few checks before you shareeach one undoable
10 people · 5 themes

People: snapshots and transcripts

Each person’s story in moments, what’s in their way and why it matters, every line resting on the passages they said.

  • Their story, in momentshow each one felt
  • Every line Nala wrote about themfrom their passages
  • The transcript, turn by turnevery passage coded
  • You choose who countsnames masked
P7 · 3 lines · 8 passages

The case: a decision and its evidence

A doc in three parts, what you heard, what you decided and what you’ll do, where every claim brings its evidence and the margin says when one doesn’t.

  • Type # for a theme, @ for a personthey bring their counts
  • The margin checks every claimand never blocks
  • Nala drafts, you decidekeep or dismiss
  • Comment, share and publishread by anyone
5 claims ·unchecked

Your AI, connected to Nala

Connect Nala to Claude, ChatGPT or another AI. Ask about your interviews and get counts with links to the quotes; anything it writes waits for a person to keep.

  • Connect in a minutea key for each member
  • Answers come back countedwith a link to each quote
  • Suggestions, never editsan editor keeps or dismisses
  • Every claim checkedno quotes, and it says so
Your AI ⇄ Nala ·idle

What holds on every page. The same questions for every passage, every claim keeping its quotes, and the same answer every time.

4 layers ·one clause

The same questions, every passage

Is it evidence, what kind, how did it feel, which theme? No text-writing AI judges a passage or counts a person.

1 claim · 4 quotes

Claims you can open

A count in the Brief, a line about a person, a sentence in the case: each one opens to the quotes it rests on.

Run 1 · run 2 ·counting

The same answer twice

Every judgment is a fixed question with a typed answer, kept for its input. Run the study again and it reads the same.

How Nala reads an interview

Bring this week’s interviews.

Cancel any time, and ask for a refund within 14 days of any payment. Your data stays in London.