Case study

2027 Polls

Poll comparator for the 2027 presidential election.

Weighted average of published polls, with margin of error. No forecast.

Poll comparator (in French)
Edition of October 20, 2026
27 polls included, 7 active pollsters out of 10 tracked. Page in French.
Technologies and topics
  • Polls
  • Weighted average
  • 2027 presidential election
  • AI agents

The project

A comparator of voting-intention polls for the 2027 French presidential election. It calculates a weighted average of published polls, with a margin of error. It makes no forecast.

What the page shows

  • A ranking of the candidates, under two hypotheses of candidacies.
  • A page for each candidate.
  • Second-round matchups.
  • Month-by-month trends, for each hypothesis.
  • An explanation of how to read the score and how it is calculated.

How I work

I define the product and I decide; AI agents write the code. The code is written by AI agents; it is not hand-written code.

Current status

Edition of October 20, 2026: 27 polls included, 7 active pollsters out of 10 tracked, most recent poll from October 19, 2026. The page is in French only.

Other case studies

Cadra TV The media player with a cinema design. Trame A family app that brings together a household's calendar, groceries, tasks, meals, notes and expenses. Nashira News Association media outlet created in February 2020.