
Yurumi: the 56 seconds nobody was measuring
The question endpoint took 56 seconds and still answered NOT FOUND, while the deterministic path answered in 3 seconds with the right source. Four independent causes, each measured before being cut.
15 posts

The question endpoint took 56 seconds and still answered NOT FOUND, while the deterministic path answered in 3 seconds with the right source. Four independent causes, each measured before being cut.

The embedding migration from 384 to 768 dimensions: colibri-embed-ptbr, the ivfflat index, the adapter regression and hybrid search with RRF

Yurumi's skill search hit concrete queries and missed vague ones, because the index read a Windows-only directory, had no dedup and returned the file path where the description should be. The index went from 197 to 733 skills.

From the first iMasters scraper to a Knowledge OS with hybrid RAG, a 42-tool ReAct Agent and an MCP server — how Arachne evolved.

How Arachne evolved from a scraper into a personal knowledge platform with zero‑port infra, hybrid search and bot generation.

Capivara's chat started answering without sources: HTTP 200, a polished sentence, zero memory. Restarting the embedding service looked like the obvious fix — and was precisely what kept it from coming back. Underneath, two defects from the same family: a process that only opens its port after waking up, and a fallback that doesn't know how to undo itself.

Before swapping Arachne's search engine for Yurumi, I had to prove the new one was at least as good as the old. Shadow dual-write, real traffic, overlap@10 — and the number nobody expects: 0.511.

After the bilingual jump to 0.1614 recall@1, there was a number left on the bench: how much was the cross-encoder contributing on every search? The week I measured that latency and recall are a bill you pay on every query — and the answer ended up written as a permanent flag in the config.

Three embedding models, two search systems, one homegrown benchmark: the study that started with a question about semantic similarity and ended up revealing the tradeoffs between colibri, bge and openai — and how each one found its place in the ecosystem.

Capivara kept its memories in a local ChromaDB, with its own embedder and L1-L4 layers. Yurumi kept the ecosystem's memory in a Qdrant. Two separate memories meant a brain that never talked to the rest of the body. The unification on 08/28 migrated 182/182 memories without losing a single one: hash dedup, backup before deleting, and an --apply gate that separates the plan from the execution.

Every project's AGENTS.md was the official instruction set of the ecosystem, but it lived outside the Yurumi's memory — the ingest only ran when someone remembered to run it. The fix was a signature watcher (mtime+size) that re-ingests only what changed. Along the way: a first_run that ran a full ingest every day, an os.stat that hung when WSL wedged, and the lesson of never trusting a junction.

Yurumi stored its memory with a single embedding model trained in Portuguese. It worked — until the first question in English. The fix was not swapping the model; it was running two of them side by side and fusing the rankings with RRF. After a full LoCoMo run, recall went from 0.139 to 0.1614.

The simple question 'what does my agent actually remember?' turned into a full audit of Yurumi: 32 knowledge bases, 551 documents, a Qdrant collection with over 8 thousand vectors, and a LoCoMo baseline that handed back reality in percentage form.

107 commits, 131 backend tests + 248 frontend, 2FA TOTP, local RAG with ChromaDB, R2 backup and D1 disaster recovery — the saga of turning a personal dashboard into reliable infrastructure.

777 commits, 2818 tests, hybrid RAG, ReAct Agent with 42 tools, MCP server, and the evolution from a simple scraper to a knowledge platform.