Mindvalley Stories
Case Study · AI Mastery

161 water-damaged books turned into an insurance-ready inventory with AI

A custom GPT and a calm assembly-line workflow turned 161 water-damaged books into a complete, insurance-ready spreadsheet.

Faced with a pile of water-damaged books and an insurance claim to prepare, this student did not freeze. They designed a custom GPT, planned a gentle step-by-step workflow, and documented all 161 books. It is a quiet, practical story about using AI to turn overwhelm into something you can finish.

David Andruss
A book owner with a large music and literature library who needed to document water damage for insurance.
Damaged Books: AI-assisted insurance inventory

Before

Water had damaged a large collection of books, many of them music scores, biographies and titles in several languages. Every one needed to be identified, priced and recorded for an insurance claim. Done by hand, it was the kind of task that feels too big and too stressful to even begin.

“Planned the workflow with ChatGPT first (“assembly line” approach: one book → 2 photos → next book) to minimize stress and friction”

In their own words

The turning point

In the program they learned to plan with AI before acting. They mapped out the job with ChatGPT first, settling on an assembly line: one book, two photos, next book. Then they built a custom GPT with extraction rules written together, so each batch of photos came back as clean rows ready for the spreadsheet.

After

They photographed every book's cover and spine, worked in batches of about five books, and pasted the results into a master sheet. A one-command Terminal script resized the images into a separate folder so the originals stayed safe. Several research passes filled in ISBNs and replacement costs, with unresolved items flagged and cleaned up at the end.

What it lets people do

The workflow lets anyone facing a damaged collection turn photos into a structured, traceable inventory without typing each entry by hand. The custom GPT pulls titles, authors, languages and publishers straight from images, ready for a spreadsheet.

What they learned

They saw that AI can reduce overwhelm by breaking a stressful task into small, repeatable steps. Simple automation removed the biggest bottleneck, and working in iterative passes with versioned files made the result more accurate and easier to trace.

“AI can reduce overwhelm by turning a stressful task into small, repeatable steps”

In their own words

See the work

Book photos turned into inventory rows by the custom GPT
First book: three photos become entry 001
Later batch: scores, philosophy and yoga books catalogued
Setting up spreadsheet column headers with ChatGPT

Damaged Books: AI-assisted insurance inventory was built during AI Mastery.

Read the full submission
Achievements - Documented 161 water-damaged books (2 photos per book: cover + spine) and built a complete inventory spreadsheet - Added ISBNs and “replacement cost” prices through multiple research passes, producing an insurance-ready Excel file Tools - ChatGPT + a Custom GPT (designed with ChatGPT; extraction rules written together) - macOS Terminal (one-command batch image resizing into a separate “compressed” folder to preserve originals) - Google Sheets / Excel Methodology - Planned the workflow with ChatGPT first (“assembly line” approach: one book → 2 photos → next book) to minimize stress and friction - Processed in batches (e.g., 5 books / 10 images): upload → extract structured fields → paste into the master sheet - Filled missing ISBNs/prices iteratively in several runs, flagging unresolved items and cleaning them up at the end Learnings - AI can reduce overwhelm by turning a stressful task into small, repeatable steps - Simple automation (batch resizing) removes major bottlenecks without losing readability - Iterative passes + versioned files improve accuracy and traceability Relevant URLs - https://chatgpt.com/g/g-6972624877cc81918c6b7aecbec4acbe-damaged-books The project is outlined in full in the uploaded Word file.

Written 15 February 2026 Submitted by the storyteller

Every story is written by a Mindvalley student or by someone our teachers have worked with. How we collect and check stories

More from AI Mastery

Damaged Books: AI-assisted insurance inventory: a AI Mastery case study | Mindvalley Stories