EDA / BIT

2025 challenges — Zurich

5_GOV_2
Achieving Legal Consistency – Swiss Law RAG

How can we guarantee that AI-generated answers always align with Swiss federal law?

Problem

Our in-house OpenWebUI instance already uses a hybrid RAG pipeline (BM25 + embeddings, re-ranked with CrossEncoder and tunable relevance thresholds) to query internal documents. While this yields solid results, it is not yet robust enough for the scale and structure of Swiss legislation. A richer knowledge representation and stronger grounding are needed to avoid hallucinations and improve trust.

Why hack?

• Real-World Impact – Your prototype can land directly inside the Swiss Federal Administration and a premium Dataset. • Full corpus of Swiss law.

6_GOV_3
Don't walk straight into the trap

Can you build an AI-based tool that prevents users from buying products that are prohibited for sale in or import to Switzerland?

Problem

Imagine buying a gadget online - only to discover it's illegal in your country (Switzerland). Swiss law for example only permits class 1 laser pointers, with all others banned from import, sale, or use. Your task in this hackathon project will be to use your imagination to build a smart, AI-based, end-2-end tool that instantly checks the legality of e-commerce products, helping everyone - from travelers to online shoppers - to stay safe and compliant.

Why hack?

•Help customers avoid accidental illegal activities •Use AI to make Switzerland a safer place

7_GOV_1
Humanitarian Intelligence: A dialog system to save lives and reduce suffering through smart aid

How can we build a multilingual, intelligent AI Assistant that empowers humanitarian actors to access, compare, and learn from complex project data—instantly and meaningfully?

Problem

This challenge invites you to prototype an internal AI application for the Swiss Humanitarian Aid (SHA). Your task is to design a conversational interface that helps staff explore large volumes of project data, evaluations, and reports – making knowledge accessible, actionable, and comparable. But this goes beyond answering questions. Your assistant should help contextualize SHA’s work by drawing comparisons with other international cooperation actors (e.g., OECD DAC, ReliefWeb). The goal: strategic insight, institutional learning, and real-world impact.

Why hack?

Because your code can save lives. Humanitarian aid is under pressure—financially, logistically, and politically. Smarter tools mean better decisions, faster responses, and more impact. This is your chance to apply cutting-edge AI to a real-world challenge with global relevance. You’ll be shaping how Swiss Humanitarian Aid works in the future.

8_GOV_4
VoiceTrust