Dictate family census records into a spreadsheet you can correct
At a glance
Genealogist Dana Leeds turned spoken census transcriptions into a CSV and resolved identity inconsistencies in Excel. This workflow is listed under Family History, Census Records, Spreadsheets.
- Agents / tools
- ChatGPT, Ancestry, Text editor, Microsoft Excel
SourceBuilding a tool for these workflows? See weekly rates.
What they did
1) Have access to the original census images, ChatGPT's mobile dictation feature, a plain-text editor and Excel. Read the census information from Ancestry into the app by speaking it aloud. 2) Ask ChatGPT to organize the transcription into columns for year, name, age, birthplace, occupation, real estate and personal estate. 3) Request CSV output comparing people across census years, using name and age for a simpler comparison. Copy the output into a text editor and save it with a .csv extension. 4) Import that file into Excel, reorder rows into the family hierarchy, add relationships and correct identities against the records. Leeds merged entries for one person whose name appeared as Julia, Judea and July.
How it went
The author produced a census comparison spreadsheet and manually reorganized and corrected it in Excel.
Worth knowing
The initial AI table omitted relationships; the author added that column during the spreadsheet cleanup.