Project · Energy Data Hackdays

ETH Swiss Energy Scenarios

Switzerland’s energy future, answered in your own language.

6AI agents, each an expert in one part
100+data tables and reports from Energy Perspectives 2050+
4languages: English, German, French, Italian
The problem

Switzerland’s official energy scenarios hold the answers on the path to net zero, but they sit in dozens of data tables and long technical reports. Citizens, journalists and policymakers cannot read them quickly, and often not in their own language.

This system lets anyone ask a question in plain words and get an answer at their level, drawn from the official data.

The idea
1AskA question in English, German, French or Italian.
2RouteAn orchestrator sends it to the agents who know that part.
3AnswerData, scenarios, reports and policy are combined into one answer.
4AdaptShaped for a citizen, a journalist, a student or a policymaker.
What was built
Built with
PythonOpenAI GPT-4Multi-agentPandasPyPDF2PlotlyStreamlitClaude Code

Built at the Energy Data Hackdays, part of Swiss {ai} Weeks, on a challenge set by ETH Zurich’s Energy Science Center.