Make a father's handwritten crop ledger, kept since 1971, searchable with ChatGPT
- Posted
- by OpenAI
- Last checked
- Method:
- How we verify
Evidence: Company case studyA write-up published by the company that makes the AI tool or uses it.
Context: From OpenAI's own small-business stories page; OpenAI makes ChatGPT.
Outcome
At Sharp & Sharp Certified Seed in Allendale, South Carolina, Rachael Sharp used ChatGPT to digitize the farm's handwritten crop ledger, kept by Sharp's father since 1971, and make it searchable. 'Every single crop we've ever planted is now accessible,' Sharp says, with past planting dates, yields and practices in one place. In the field, Sharp uses voice mode to log loads, and when soybeans looked stressed, Sharp photographed them and asked ChatGPT what was wrong.
AI tool
Prompt Recreated prompt
The source didn't share its exact prompt, so we wrote this from what it describes.
Here are scanned pages from my father's handwritten crop ledger. Transcribe them into a table with the year, field, crop, planting date, yield and notes, so I can search past seasons.
Try this in:
Sends the Recreated prompt shown above. ChatGPT and Grok open with it filled in, and ChatGPT may send it straight away. Gemini opens empty, so we copy the prompt for you to paste. Every button also copies it.
Opens in a new tab. We don't see what you send.
These links are only a shortcut. We're not affiliated with OpenAI, Google or xAI, and your results may not match the source.
Optional stronger version
Our rewrite with extra checks. The poster didn't test it.
Before you start, ask me 1–3 short questions about the details below and wait for my answers. Once I've answered, confirm the plan in one line and run the whole task start to finish. Goal: Turn handwritten records into a clean, searchable table without losing or changing anything. Questions to ask me first: - What do the records cover, and what columns do I want (for example year, field, crop, date, yield, notes)? - How many pages are there, and are they photos or scans? - Are there abbreviations or field names I should explain first? Inputs (fill in or confirm): - Page images: [PAGES] - Columns I want: [COLUMNS] - Abbreviations and field names: [KEY] Rules: - Transcribe exactly what's written; don't correct, convert or fill in numbers. - Keep a page reference for every row so I can check it against the original. - Mark anything you can't read as [UNCLEAR] with your best reading in brackets. - Keep units as written and note them in a separate column. - Keep the original images; this table is a copy, not a replacement. Checkpoint: Transcribe the first two pages and wait for me to check them before doing the rest. If you get stuck: If the layout changes between years, ask me how to map it to the columns rather than guessing. Verify before you finish: Spot-check totals or repeated entries against the images and list every [UNCLEAR] item for me to review. Output: A spreadsheet-ready table (CSV) with page references, plus a list of unclear entries. Done when: Every page is transcribed, every row points back to its page, and I can search past seasons.
Source
https://openai.com/index/small-business-stories/
Related cases
The poster says they asked Grok to graph Iowa traffic deaths per month for the last five years, then California, to check whether a roadside death counter looked low; they shared the charts.
Grok · Data · Source date:
- Build a support data pipeline and dashboard in under an hour with DevinRecreated promptEvidence: Company case study
At Eight Sleep, a support manager who used to need the data team for dashboards built an end-to-end pipeline and dashboard themselves with Devin in under an hour, according to Devin's customer story.
Poster says:Under an hour
Devin · Data · Source date:
- Set up a Google Analytics dashboard with Claude in ChromeImproved promptEvidence: Described by posterFailure
Claude in Chrome spent 45 minutes building a Google Analytics dashboard, then couldn't work out how to drag one element, so it undid all of its work and told the poster it couldn't do it.
Claude in Chrome · Data · Source date: