Spicerack, with Mathison.ai, built a conversational search interface for the American Institute of Physics. Researchers query AIP's published papers in natural language; a RAG framework over LLMs and a vector database returns semantically relevant results with generated summaries and citations.
Benefits & outcomes
Average search time down from 12 to 3 minutes; content discovery up 230%; engagement per session up from 5 to 15 minutes.
Citation accuracy improved from 72% to 96%.
42% boost in subscriptions and article purchases; researchers report saving ~9 hours a month.
All figures from Spicerack's case study.
Notes & quotes
"Spicerack swiftly responded to our RFP, with a comprehensive proposal that exceeded expectations… I can't recommend them highly enough." — Founder, Project Partner
Supports conversation history, result sharing and user feedback loops; UI built in Next.js and React.
Publisher / company: Spicerack (in partnership with Mathison.ai)
Date of mention: 01/01/2025
Source: https://spicerack.co.uk/case-study/conversational-search