Step 2 Results: Quote-Backed Hotel Evidence Extraction
Step 1 ended with four frozen, hash-verified source texts for Omni Mount Washington Resort & Spa. Step 2 tested whether a local model could turn those files into…
Step 1 ended with four frozen, hash-verified source texts for Omni Mount Washington Resort & Spa. Step 2 tested whether a local model could turn those files into…
Hotel source acquisition now has a production boundary. A controller can start with a stable property identity, ask a local model for bounded searches and source selections, retrieve…
Updated 12-hotel qualification benchmark showing how a bounded Firecrawl fallback changed extraction readiness for Gemma 4 31B and Qwen 3.5 9B.
Updated 24-run ledger including bounded Firecrawl recovery, retained sources, hashes, failures, and final package status.
A 28-route benchmark starting with only a hotel name and full address: models planned searches, selected websites, and faced deterministic retrieval validation.
Every query plan, selected website, raw response, scoring receipt, and retrieval result from the controlled hotel source-discovery benchmark.
A grid comparing every installed non-coding, non-vision text model on the bounded hotel source-selection task, plus paired local-versus-remote speed trials.
Exact prompt, candidate catalog, scoring receipts, telemetry, and raw output from 28 local and remote text-model runs.
Executed first-step source acquisition for the Skokie DoubleTree, with the exact frozen source package the next no-web extraction card would receive and full failure/recovery disclosure.
The exact first-wave local-model discovery cards for three hotel classes, including complete prompts, limits, dependencies, outputs, receipts, and stop conditions.