Humanitarian reports, as data
Every week your team reads dozens of PDF reports to find the same few numbers. We set up a pipeline that reads them for you. It collects new reports, pulls out every figure with AI, checks each one against the report's own words, and keeps them in a table you can search, chart and share.
For NGOs, UN agencies, clusters, researchers and journalists who track displacement, casualties, access or funding across many reports, in Arabic or English.
How it works
- Collect. New reports arrive each morning from ReliefWeb, your partners' emails, or a shared folder.
- Extract. An AI model reads each report and writes down every figure, its place and date, and the exact sentence it came from.
- Check. Each sentence is searched for in the report. A figure we can't find word for word is marked unverified, so a guess never passes as fact.
- Use. The figures land in BigQuery or a file you own, with a live tracker and a CSV export.
One sentence, as data
“Since the beginning of 2026, approximately 1,790 Palestinians have been displaced by demolitions for lack of Israeli-issued building permits across the West Bank.”
OCHA, Humanitarian Situation Report, 22 October 2026
- Indicator
- People displaced
- Value
- 1,790
- Place
- West Bank
- Period
- 1 January to 22 October 2026
- Checked against the report
- Found word for word
What you get
- A pipeline for the sources and indicators you follow, running every day.
- Your choice of model. Gemini for the best reading of long Arabic and English reports, or an open-source model on your own computer, so sensitive documents never leave it.
- A live tracker and a table in BigQuery, Google Sheets or a file, with a written definition of every indicator.
- The code, open on GitHub, so your team can keep it running without us.
Priced per project, after a free 20-minute call.