Skip to content

run_2026-07-30_nightly

Nightly · v3.2 · 2026-07-30

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 (424 items).

Pass rate

78.5%

+1.4 pts(improvement)vs previous nightly333 of 424 samples passed

Faithfulness

0.86

0.00(no change)vs previous nightly

Answer correctness

0.81

−0.01(regression)vs previous nightly

Context precision

0.70

+0.01(improvement)vs previous nightly

Context recall

0.74

−0.02(regression)vs previous nightly

Latency p95

3.3s

−0.3s(improvement)vs previous nightly

Failures by category

91 failed samples · largest share: unsupported

Share of failed samples by category, compared with Nightly · v3.2 · 2026-07-29
CategorySharevs reference
Unsupported claims24.2%0.0 pts
Incorrect factual answers18.7%−0.3 pts
Missing context23.1%−0.1 pts
Retrieval mismatch24.2%0.0 pts
Instruction-following failures6.6%+0.3 pts
Other3.3%+0.1 pts

Run details

Dataset
support-golden-v4 v4.6 · 424 items
Generator
Claude Sonnet 5.5
Retriever
bge-m3 · dense
Prompt
support-prompt v5
Judge
LLM judge · Claude Opus 5.5 · rubric v3
Started
Jul 30, 02:00 UTC
Duration
40m 2s
Triggered by
Scheduler
Tokens / request
3,870 in · 230 out
  • nightly
  • golden-slice

All metrics

Against Nightly · v3.2 · 2026-07-29

All metrics for Nightly · v3.2 · 2026-07-30
MetricThis runReferenceChange
Pass rate78.5%77.1%+1.4 pts
Faithfulness0.860.860.00
Answer correctness0.810.82−0.01
Instruction adherence0.860.860.00
Context precision0.700.69+0.01
Context recall0.740.76−0.02
Recall@50.780.780.00
MRR0.630.65−0.02
NDCG@100.730.730.00
Latency p501.8s1.6s+0.1s
Latency p953.3s3.6s−0.3s
Cost per request$0.0045$0.0043+0.0002

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
78.5%
0.860.740.783.3sThis 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 · 2026-07-30

How long does it take to port my number from another operator?

Number porting

FailIncorrectjudge 0.15

Why was I charged twice this month?

Billing & payments

FailUnsupportedjudge 0.35

My phone shows 4G but mobile data does not work after the iOS update.

Device troubleshooting

Passjudge 0.93

How do I get a copy of the personal data Harbor holds about me?

Data privacy requests

Passjudge 0.91

What is the late fee if I pay my bill after the due date?

Billing & payments

Passjudge 0.92

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

Number porting

Passjudge 0.88

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

Promotions

Passjudge 0.90

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

Returns & warranty

Passjudge 0.92

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

Billing & payments

Passjudge 0.88

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

eSIM & activation

Passjudge 0.98