Skip to content
← All templates
🎤

Customer Interview Repository

A shape for the interviews that matter. Save the raw notes, tag the themes, and let Claude or ChatGPT do the synthesis from real material. Stop re-reading the same transcripts.

Requires an AI connected to your Hjarni account via MCP.

https://hjarni.com/templates/customer-interview-repository

Copy this URL and paste it into Claude or ChatGPT to install the template.

How to use

  1. 1 Share this page. Paste this URL into Claude or ChatGPT. Your AI reads the template definition and installs it.
  2. 2 Folders, tags, and instructions appear. Your AI creates the full structure in your Hjarni account, ready to use.
  3. 3 Start adding notes. The AI instructions guide your AI on where to put things and how to organize them.

A shape for the interviews that matter. Save the raw notes, tag the themes, and let Claude or ChatGPT do the synthesis from real material.

Thirty interviews in, no one on the team remembers exactly what each customer said. The notes are saved, somewhere. The themes live in someone's head. This template gives those interviews a structure your AI can actually use.

Real quotes, not invented customer voice

Claude and ChatGPT are useful for synthesis, but only when they are grounded in the actual interviews. This template keeps raw notes, themes, and quotes separate, with folder-level instructions that tell the AI to cite the source interview every time.

When you ask "what did our last ten customers say about pricing", you get answers tied to specific notes by name and date, not invented social proof.

A sample interview note format

Every interview note follows the same shape. The starter note in the Interviews folder is exactly this skeleton, ready to copy.

# Interview - Sarah - 2026-05-16

Role:
Company size:
Current workflow:
Pain points:
Quotes:
Feature requests:
Willingness to pay:
Follow-up questions:

A weekly synthesis workflow

  1. After each interview, drop the notes into the Interviews folder.
  2. Once a week, ask Claude or ChatGPT to scan recent interviews for new themes.
  3. Save confirmed themes and reusable quotes back into the Themes and Quotes folders.
  4. Use the repository when writing copy, roadmap notes, or investor updates.

Six months in, you have a memory of your customers that survives team changes and that your AI can actually read.

A real example

You are writing a launch email. You ask ChatGPT, "Pull three real quotes from customers about how they currently solve this problem." ChatGPT reads the Quotes folder, returns three quotes with names and dates. You paste them in.

A week later, an investor asks "what are the top three pain points your customers raise". You ask Claude. Claude reads the interviews and themes tagged #pain-point and answers with frequencies and citations. Your call notes become your data.

Common questions

Common questions

Does Hjarni record or transcribe customer calls?

No. Hjarni stores the notes you save. Bring notes from your call recorder, transcript tool, or manual notes.

Will the AI invent customer quotes?

It can if you ask it to write from memory without notes. This template is designed to avoid that by storing real quotes and instructing Claude or ChatGPT to cite the source interview.

Where do transcripts live?

In the Interviews folder. Long transcripts can also be linked as Hjarni file attachments on the interview note.

Can I share this with my team?

Yes. Install in a team space so every researcher's Claude or ChatGPT reads from the same repository. See Hjarni for Teams.

Related pages

Templates overview | Hjarni for Founders | Custom AI instructions per folder | Hjarni for Teams | Hjarni vs ChatGPT Memory

Structure

Tags

pricing onboarding feature-gap competitor-mention willingness-to-pay pain-point

Folders

Customer Interview Repository
The shared memory of every customer conversation. Real quotes, named sources, never invented.
Has AI instructions
4 starter notes
Interviews
One note per interview. Raw notes, attendee, date, role, company size.
Has AI instructions
Themes
Recurring patterns extracted across interviews.
Has AI instructions
Quotes
Single-line quotes you want to reuse in writing and pitches.
Has AI instructions
Open Questions
Questions you have not asked yet, or have not figured out.
Has AI instructions

For your AI

Share this page with your AI. It reads the definition below, creates the folders, instructions, tags, and starter notes in your account.

Show template definition

Install steps for AI agents:

  1. Check existing tags with tags-list. Only create missing ones with tags-create.
  2. Create containers top-down using containers-create, noting the returned IDs. Use parent_id to build the hierarchy.
  3. For each container with llm_instructions, call instructions-update with level: "container" and the container's ID.
  4. Create any seed notes using notes-create, placing them in the correct container by ID. Use container_path to resolve which container.
  5. Discuss any customizations with the user before or after installing.
---
name: Customer Interview Repository
description: 'A shape for the interviews that matter. Save the raw notes, tag the
  themes, and let Claude or ChatGPT do the synthesis from real material. Stop re-reading
  the same transcripts.

  '
tags:
- pricing
- onboarding
- feature-gap
- competitor-mention
- willingness-to-pay
- pain-point
containers:
- name: Customer Interview Repository
  description: The shared memory of every customer conversation. Real quotes, named
    sources, never invented.
  llm_instructions: |
    This is the customer interview repository. The goal is to give Claude and ChatGPT real material to synthesize from, never an excuse to invent customer voice.
    - Never fabricate quotes, names, dates, or themes. If the user asks for material that the repository does not contain, say so plainly.
    - Cite the source interview by first name and date every time you quote a customer.
    - When asked about pain points, themes, or willingness to pay, search the Interviews and Themes folders first. Pull supporting quotes from the Quotes folder when they exist.
    - Suggest filing new themes back into the Themes folder when a recurring pattern shows up across interviews.
    - This is a knowledge base for notes, not a transcription or recording tool.
  children:
  - name: Interviews
    description: One note per interview. Raw notes, attendee, date, role, company
      size.
    llm_instructions: |
      Use this folder for raw interview notes.
      - One note per interview. Title format: "Interview - <First Name> - YYYY-MM-DD".
      - Use the shipped skeleton: Role, Company size, Current workflow, Pain points, Quotes, Feature requests, Willingness to pay, Follow-up questions.
      - Always cite the interviewee by first name and date when referencing this note. Never invent quotes that are not in the body.
      - If the user shares a transcript or recording link, save it in source_url or attach it to the note. Do not transcribe.
      - Tag pain points with "pain-point". Other recurring topics use the shipped tags.
  - name: Themes
    description: Recurring patterns extracted across interviews.
    llm_instructions: |
      Use this folder for synthesized themes that show up across multiple interviews.
      - One note per theme. Title is a short noun phrase, not a sentence.
      - When asked about a theme, return the supporting interview notes by first name and date. If only one interview mentions it, say so plainly.
      - Update theme notes in place when new supporting interviews arrive. Do not append in a journal style.
      - Cross-link to the source interviews using wiki-links.
  - name: Quotes
    description: Single-line quotes you want to reuse in writing and pitches.
    llm_instructions: |
      Use this folder for reusable customer language.
      - One quote per note, with attribution (first name, date) and a short link back to the source interview.
      - When drafting marketing copy or external content, prefer real quotes from this folder over invented ones.
      - Never invent or paraphrase a quote into a more polished version. Quote the words as said.
  - name: Open Questions
    description: Questions you have not asked yet, or have not figured out.
    llm_instructions: |
      Use this folder for unresolved customer research questions.
      - Keep each note short: the question, why it matters, and any partial answers.
      - When a question is answered by a new interview, fold the answer into the relevant Theme note and resolve the question.
      - When asked something open, check here first and say so plainly if it is unresolved.
  notes:
  - title: Interview - Sarah - 2026-05-16
    body: |
      A starter note showing the shape. Replace it with a real interview.

      **Role:** Head of Product, mid-stage SaaS
      **Company size:** 18 people

      **Current workflow:**
      Uses a mix of Notion docs, Slack, and a private folder of voice memos to track customer feedback. Synthesis happens in her head before each sprint planning.

      **Pain points:**
      - Re-reading the same notes every quarter.
      - Losing track of which customer asked for what.
      - Hard to share context with the rest of the product team without rewriting it.

      **Quotes:**
      - "I write good notes. I just never read them again."
      - "Every quarter I rebuild the same picture from scratch."

      **Feature requests:**
      - A way to search interviews by theme rather than by date.
      - Something her teammates' AI can also read.

      **Willingness to pay:**
      Would consider a team plan if it removed the synthesis bottleneck.

      **Follow-up questions:**
      - Ask which themes she is currently tracking by memory.

      This is a starter note. Replace it with a real interview.
    tags:
    - pain-point
    - feature-gap
    container_path: Customer Interview Repository > Interviews
  - title: 'Theme: Synthesis bottleneck'
    body: |
      A starter theme note showing the shape.

      ## Summary
      Multiple founders describe a pattern where interview notes pile up faster than they get synthesized. The notes exist; the memory does not.

      ## Supporting interviews
      - [[Interview - Sarah - 2026-05-16]]

      ## Open questions
      - Does the bottleneck shrink when AI does the synthesis, or just shift?

      This is a starter note. Replace it with a real recurring theme.
    tags:
    - pain-point
    container_path: Customer Interview Repository > Themes
  - title: 'Quote: Sarah on rebuilding context'
    body: |
      > "Every quarter I rebuild the same picture from scratch."

      Sarah, Head of Product at a mid-stage SaaS. Recorded 2026-05-16. Source: [[Interview - Sarah - 2026-05-16]].

      This is a starter note. Replace it with a real quote.
    tags:
    - pain-point
    container_path: Customer Interview Repository > Quotes
  - title: 'Open question: pricing anchor'
    body: |
      ## Question
      What price point do founders use as their internal anchor when comparing knowledge base tools?

      ## Why it matters
      Pricing claims in interviews often reference a competitor without naming the price.

      ## What we know
      - Several interviews mention "Notion-like" pricing without a number.

      ## What we do not know
      - Whether the anchor is per-seat or flat per team.

      This is a starter note. Replace it with a real open question.
    tags:
    - pricing
    container_path: Customer Interview Repository > Open Questions

Write once. You both remember.

Free to start. No credit card required.

Give your AI a memory

Works with Claude and ChatGPT today. Gemini coming soon.