Data plane · active learning (B6)
Training studio
Claude labels → ML.NET serves at scale → low-confidence cases feed the loopM1 · Condition classifierserving
new · like-new · good · fair · poorv1.7
Accuracy94.2%
Labels18.4k
Retrain 24h+412 this week
M2 · Feature extractorserving
sealed · gatefold · 180g · OEM · with_casev2.1
Accuracy91.5%
Labels22.9k
Retrain 24h+536 this week
M3 · Product matcherserving
cross-venue identity · gray-zone pairsv1.3
Accuracy88.9%
Labels9.1k
Retrain 24h+287 this week
M4 · Brand / Categoryserving
brand + category from titlev3.0
Accuracy96.1%
Labels31.6k
Retrain 24h+624 this week
Active-learning loop
looping · hot-swap
Sample low-confidence
→
Claude labels
→
Grow seed set
→
Retrain
→
Serve
💡
The ML.NET classifiers serve at scale; only cases below the confidence threshold escalate to Claude for labels (ConditionLabel / FeatureSynonym / ProductMatchLabel). New labels grow the seed set, the model retrains every 24h and hot-swaps in.
| Type | Example title | Label | Decided by | When |
|---|---|---|---|---|
| ConditionLabel | Sony WH-1000XM4 — open box, tested | like-new | 🤖 Claude | 2m ago |
| FeatureSynonym | Miles Davis Kind of Blue 180 gram audiophile | 180g | 🤖 Claude | 14m ago |
| ProductMatchLabel | "WH1000XM4" ↔ "WH-1000 XM4 headphones" | match | 🤖 Claude | 31m ago |
| ConditionLabel | Canon G7X II - some scratches, works great | good | 🤖 Claude | 48m ago |
| FeatureSynonym | Rolex Sub w/ box and papers full set | with_case | 👤 Operator | 1h ago |
| ProductMatchLabel | "Prusa i3 MK3S" ↔ "Original Prusa MK3S+" | no-match | 👤 Operator | 2h ago |
Training studio · Phase 3 — self-serve site config
PlannedA non-engineer adds a new site config without code: LLM-assisted JSON mapping plus point-and-click DOM annotation, with a Tier-2 CSS / XPath fallback when structured feeds aren't available. Extends the field-mapping table above to net-new sources.