Vectorless RAG: retrieve by reasoning, not by similarity
Why "chunk, embed, top-k" struggles on long financial reports, and how a table-of-contents tree that an LLM navigates can find the right pages instead.
ragretrievalagents
Read →blog
Learning in public
Notes on concepts I've picked up from papers, docs, talks and building things. Every post links the sources I learned from. RSS feed
MCP in 10 minutes: one plug for every tool
What the Model Context Protocol actually standardises: hosts, clients and servers, the three things a server can offer, and what changed in the stateless 2026 spec.
mcpagentsllm
Read →