Trieve Blog

Using a LLM-as-a-judge for hallucinations is slow and imprecise relative to simple NER. We share how we solved hallucination detection at Trieve.

Unmetered In-VPC Embeddings and Rerankers at Ridiculously Low Latency

Modern, configurable drop-in search and RAG component

PGVector offers infrastructure simplicity at the cost of missing some key features desireable in search solutions. We explain what those are in this blog.

Instructions for self-hosting Trieve on a VPS using docker-compose. You'll be able to set up Trieve on a Hetzner server which comes with semantic and hybrid search, SPLADE fulltext search, re-ranker models, RAG AI Chat, recommendations, and analytics.

With heavyweight backers Root Ventures and Y Combinator, Trieve is building the best solution for every developer building an AI application.