Case Study: Allspice boosts ingredient matching accuracy with Pinecone

Allspice boosts ingredient matching from 20% to 97% with Pinecone

Allspice, a food technology company building a kitchen operating system, struggled with the inherent messiness of recipe and ingredient data. Their traditional text search could not reliably match ingredient descriptions, which was blocking a core product feature and key revenue stream. They turned to Pinecone to implement a dedicated semantic search solution to solve this challenge.

By implementing Pinecone's vector database, Allspice created a semantic layer that bridges messy language and structured data. This solution improved their ingredient matching accuracy from 20% to 97%, enabling the launch of their recipe importing feature. The use of Pinecone has expanded across their platform, improving user satisfaction and generating new revenue opportunities for their publishing partners.

Allspice