Skip to content

run_2026-08-01_nightly

Nightly · v3.2 · 2026-08-01

Claude Sonnet 5.5 · bge-m3 · dense · support-prompt v5 on support-golden-v4 v4.6

CompletedNightlysupport-assistant v3.2v2 · 512 tokens / 64 overlap
Compare

Stratified slice of support-golden-v4 (420 items).

Pass rate

78.6%

+1.9 pts(improvement)vs previous nightly330 of 420 samples passed

Faithfulness

0.85

−0.01(regression)vs previous nightly

Answer correctness

0.81

0.00(no change)vs previous nightly

Context precision

0.70

+0.01(improvement)vs previous nightly

Context recall

0.76

+0.01(improvement)vs previous nightly

Latency p95

3.4s

−0.1s(improvement)vs previous nightly

Failures by category

90 failed samples · largest share: unsupported

Share of failed samples by category, compared with Nightly · v3.2 · 2026-07-31
CategorySharevs reference
Unsupported claims24.4%+1.0 pts
Incorrect factual answers18.9%−0.3 pts
Missing context23.3%−0.1 pts
Retrieval mismatch23.3%−0.1 pts
Instruction-following failures6.7%−0.8 pts
Other3.3%+0.1 pts

Run details

Dataset
support-golden-v4 v4.6 · 420 items
Generator
Claude Sonnet 5.5
Retriever
bge-m3 · dense
Prompt
support-prompt v5
Judge
LLM judge · Claude Opus 5.5 · rubric v3
Started
Aug 1, 02:00 UTC
Duration
45m 8s
Triggered by
Scheduler
Tokens / request
3,870 in · 230 out
  • nightly
  • golden-slice

All metrics

Against Nightly · v3.2 · 2026-07-31

All metrics for Nightly · v3.2 · 2026-08-01
MetricThis runReferenceChange
Pass rate78.6%76.7%+1.9 pts
Faithfulness0.850.85−0.01
Answer correctness0.810.810.00
Instruction adherence0.870.870.00
Context precision0.700.69+0.01
Context recall0.760.75+0.01
Recall@50.780.79−0.01
MRR0.640.63+0.01
NDCG@100.720.71+0.01
Latency p501.6s1.6s0.0s
Latency p953.4s3.5s−0.1s
Cost per request$0.0044$0.00440.0000

Nightly trend

Nightly runs on the support-golden-v4 slice, Jul 11 to Oct 8. This run is one point in the series.

Pass rate moved from 78.5% to 87.2% across 90 nightly runs. Configuration changes: v3.3 on Aug 4, v3.4 on Sep 8.

Across datasets

support-assistant v3.2 on every dataset in the project. Rows marked “matrix” come from the comparison matrix rather than a dedicated run.

Configuration results per dataset
DatasetItemsPass rateFaith.Ctx recallRecall@5p95Source
support-golden-v4this run1,240
78.6%
0.850.760.783.4sThis run
billing-edge-cases312Not evaluated with this configuration
roaming-and-travel188Not evaluated with this configuration
device-troubleshooting426Not evaluated with this configuration
account-security154Not evaluated with this configuration
locale-formatting96Not evaluated with this configuration
synthetic-hard-negatives540Not evaluated with this configuration

Traced samples

17 samples with full traces (4 failing, 13 passing), stratified by domain. Open a row to inspect retrieval and the judge's reasoning.

Traced samples for Nightly · v3.2 · 2026-08-01

Is there a restocking fee if I return an opened phone?

Returns & warranty

FailMissingjudge 0.13

My son is 15. Can he have his own line on my account?

Plans & upgrades

FailIncorrectjudge 0.10

Does the Fall 2026 upgrade offer work with a family plan discount?

Promotions

Passjudge 0.86

How long does it take to port my number to Harbor?

Number porting

Passjudge 0.90

My bill says 41,20 €. Is that forty-one euros or four thousand?

Billing & payments

Passjudge 0.89

Can I return a phone I bought online if I opened the box?

Returns & warranty

Passjudge 0.91

Is SIM swap protection turned on by default?

Account & security

Passjudge 0.90

My eSIM QR code says it has already been used. What should I do?

eSIM & activation

Passjudge 0.93

How much data can I use in the EU on the Unlimited Plus plan?

Roaming & travel

Passjudge 0.90

If I upgrade my plan mid-month, how is my next bill calculated?

Plans & upgrades

Passjudge 0.94