Demo Plant · Refinery North
AR

RAG Evaluation Dashboard

Benchmark metrics comparing system answers against curated expected output.

📊

Live System Benchmark Evaluation

The metrics shown below are calculated directly by executing standard benchmark Q&A pairs against the active database index. Report log: data/benchmark/results_sample.json. Last evaluated: Fri, 03 Jul 2026 16:08:18 GMT.

Benchmark Questions

40

questions.json

Top-1 Source Hit Rate

20.0%

target doc in rank 1

Top-3 Source Hit Rate

35.0%

target doc in top 3

Asset Hit Rate

100.0%

target tag resolved

Average Latency

100.6ms

embedding + retrieval

Failed Questions

26

below context threshold

Benchmark Run Items

Click on any row to inspect expected vs. actual answers

IDCategoryTarget DocTop 1Top 3Asset HitLatencyStatus
Q01equipment specpump_oem_manual.pdf300msfailed
Q02spare partspump_oem_manual.pdf96mspassed
Q03spare partspump_oem_manual.pdf95mspassed
Q04failure modepump_oem_manual.pdf97msfailed
Q05proceduresop_pump_startup.pdf97msfailed
Q06proceduresop_pump_startup.pdf108msfailed
Q07inspection findinginspection_report_q1_2025.pdf94msfailed
Q08compliance gapinspection_report_q1_2025.pdf94mspassed
Q09compliance gapinspection_report_q1_2025.pdf97msfailed
Q10compliance gapcompliance_checklist_2025.pdf96mspassed
Q11regulatorycompliance_checklist_2025.pdf92msfailed
Q12maintenance historywork_orders_clean.csv91mspassed
Q13asset timelinework_orders_clean.csv96mspassed
Q14asset lookupsop_compressor_operations.pdf95msfailed
Q15knowledge graph_relationshipcompliance_checklist_2025.pdf93mspassed
Q16failure modepump_oem_manual.pdf94msfailed
Q17equipment speccompressor_datasheet.xlsx94msfailed
Q18maintenance historywork_orders_clean.csv91mspassed
Q19failure/RCArca_findings_2025.csv94msfailed
Q20regulatorycompliance_checklist_2025.pdf94msfailed
Q21knowledge graph_relationshipwork_orders_clean.csv98mspassed
Q22citation accuracycompliance_checklist_2025.pdf100msfailed
Q23compliance gapinspection_report_q1_2025.pdf106msfailed
Q24proceduresop_compressor_operations.pdf93msfailed
Q25regulatorycompliance_checklist_2025.pdf92msfailed
Q26equipment speccompressor_datasheet.xlsx95msfailed
Q27asset timelineinspection_report_q1_2025.pdf91msfailed
Q28equipment specpump_oem_manual.pdf90mspassed
Q29proceduresop_pump_startup.pdf91msfailed
Q30multi-document reasoningcompliance_checklist_2025.pdf91mspassed
Q31multi-document reasoningcompressor_datasheet.xlsx93msfailed
Q32citation accuracypump_oem_manual.pdf90mspassed
Q33spare partscompressor_datasheet.xlsx91msfailed
Q34asset timelineinspection_report_q1_2025.pdf90msfailed
Q35asset lookupwork_orders_clean.csv92mspassed
Q36failure/RCArca_findings_2025.csv99msfailed
Q37equipment specpump_oem_manual.pdf98mspassed
Q38proceduresop_pump_startup.pdf94msfailed
Q39knowledge graph_relationshipinspection_report_q1_2025.pdf97msfailed
Q40compliance gapcompliance_checklist_2025.pdf107msfailed

Run Inspector

Detailed retrieval comparison

📋

Select any question row on the left to inspect expected answers, extracted entities, and actual grounding text results.