For UK charities · Survey & monitoring data
Make sense of your survey results just by asking
Rows of records into clear figures and charts. And none of it leaves your systems.
I build UK charities private AI assistants that answer questions from the systems they already run, without any of it leaving their control. Each one is built for one job; this one reads the survey data in your spreadsheets, whether that's programme feedback, outcomes monitoring or field records.
You ask in plain English, it answers with the figures, the tables and the charts, and every figure links back to the records it came from.
The quick way to do this today is to export the data and paste it into ChatGPT, which sends it to a US company, into logs you can't see, with no lawful basis to be there. Survey data can carry personal details, or things that have to stay confidential, and a public chatbot can also miscount, or quietly invent a pattern that nobody recorded.
The assistant does the same analysis without any of it leaving your systems, and shows its working.
There's also one that reads across all your systems at once. Below is this assistant doing its job: turning a season of survey records into figures you can put straight in a report.
What you'd ask it
These run against a made-up charity's survey data, the Fenmere Trust's wildlife monitoring log and session register, kept in a Google Sheet so I can show you without a real record on screen. Each one starts with what you're actually trying to do, then the question you'd type, then the assistant answering and citing the rows it drew from.
You're pulling this year's monitoring figures together for a report, and you want the count for one species at one site to be right, not a rough total.
You'd open the assistant and ask:
"What do our surveys show for water voles at Hartswick Fen this year?"
A funder wants the trend for a priority species over time, as a chart you can drop straight into the report.
You'd open the assistant and ask:
"How have woodlark counts changed across our reserves over the last five years?"
You want to thank your most active volunteer recorder, and you want their contact details without digging through two systems.
You'd open the assistant and ask:
"Who recorded the most surveys this year? Pull up their contact record so I can thank them."
Someone outside asks what you've recorded about a sensitive species, and you need the figures plus a clear line on what you can and can't share.
You'd open the assistant and ask:
"A local birding group has asked about stone-curlew at Fenmere. What do our surveys show this year, and what are we allowed to tell them about the location?"
These demos read records in a Google Sheet, which is just what I used for the demo. Your survey might run in Google Forms, Microsoft Forms, SurveyMonkey or Typeform, or the records might be typed in by hand, but it all lands in a spreadsheet either way. If it lives in a sheet, the same analysis works the same way.
Want an assistant like this for your own survey data, or have a question of your own in mind? Tell me what you're trying to learn from your records →
Where your survey data goes
Nowhere it shouldn't. The assistant runs in an account you own. The records, your questions and its answers are never sent to ChatGPT, Gemini, Claude or any other outside service. It reads the sheet with its own read only access, so it can count and summarise but cannot change, delete or export a record. If it tried, your systems would refuse.
And it never does the arithmetic itself. The counting, the peaks and the totalling are done by ordinary code, so the same question gives the same figure every time and it matches your spreadsheet. Every figure links back to the exact records behind it, it draws only from what the records actually say rather than inventing a pattern that reads well, and before any of it reaches your team it is tested against data where the right answers are already known.
On numbers you'll put in a funder report, that matters more than anything. Here's how I test AI assistants and what the testing catches →
Common questions
Can I use ChatGPT to analyse survey data?
You can paste it in, but you shouldn't with real data. Survey records often name people or carry detail that shouldn't be public, and the public version of ChatGPT sends that to OpenAI in the United States, keeps it in logs you can't see, and may use it to train future models. For a UK charity that's personal data leaving your control with no lawful basis to justify it.
Is it safe to put survey data into ChatGPT?
Often not. Even data that looks routine can carry sensitive detail, sometimes special category data, a respondent's personal circumstances, or a location that has to stay confidential like a protected species' nest site. Pasting it into a public tool sends it outside your charity entirely, with none of the lawful basis, contract or data protection impact assessment that would normally be required.
Can AI count survey data accurately?
Yes, if the counting is done by code rather than the model guessing. A private build runs the actual sums, peaks and cross-tabs deterministically, so the figure is the same every time and matches your spreadsheet. That's the difference between a number you can put in a funder report and one you have to check by hand afterwards.
Can it read the notes, not just the numbers?
Yes, and that's often the point. Alongside the counts it reads the free-text notes on each record, groups them and cites each back to the row it came from, so you can see how many records say a thing and read their exact words. It won't invent a pattern that isn't there or quietly bury an awkward one.
Our survey runs in Google Forms or SurveyMonkey. Can it use that?
Yes. Those all export their records to a spreadsheet, and the assistant reads the spreadsheet inside your own systems. You keep collecting exactly as you do now; the analysis just happens somewhere private instead of in a public tool.
Can the AI change or delete our survey data?
No. The build reads a copy of the records, so it can count and summarise but cannot write, delete or export. It also shows the records behind every figure, so anything it tells you can be checked against the sheet.
Got a question that isn't here? Ask me directly →
One job is just the start
Everything on this page is one assistant, built for a single job. That's deliberate: most charities prove it on the one task that's hurting, then add the next.
The most capable assistant I build reads across your CRM, your finance system, your spreadsheets and your documents at once, and answers questions that cut across all of them. One question, one plain English answer, every figure linking back to the real record it came from.
See what that looks like when it all joins up. Watch the full walkthrough →
