Explore 8 experiments across 6 themes
Interactive tool for sending the same prompt to multiple LLMs in parallel and comparing responses, latency, token usage, and cost. Uses LiteLLM + OpenRouter for model access and Langfuse for observability.
Fetches information about various WRI assets (currently datasets) and displays them visually for a user to search and explore.
Interactive tool for navigating AI provisioning decisions. Includes a constraint-based decision wizard, TCO calculator with adjustable assumptions, side-by-side method comparison, and a basic reference guide.
Client-side semantic search tool for PDF documents using browser-based embeddings. Upload a PDF, search by meaning, and see relevant pages highlighted in a visual heatmap.
Chat interface for MapLibre GL JS maps, built to test whether MCP running in a web worker is a viable architecture for browser-based LLM tool calling.
A template example showing the structure of an experiment in this repository.