Tag

AI & LLMs

Generative AI, LLMs, prompt engineering, RAG and embeddings.

2 notes

Sep 10, 20267 min read
Pinecone vs Postgres pgvector: when each makes sense
I started ReclamaAI with Pinecone. I considered migrating to pgvector. This is the honest comparison with costs, latencies, and the final decision with reasons.
AI & LLMsDatabaseReclamaAI
Apr 23, 20267 min read
Prompt engineering in production: what we learned building with LLMs
After a year iterating on the prompts behind a real product with Claude, here's what I know works and what I know is noise. No magic recipes, principles instead.
AI & LLMsReclamaAI