STAYCHARTED INSIGHTS
Business-specific AI.
Lessons from building it.
Benchmarks, engineering decisions, and what it takes to turn your examples into useful results.
Can machine learning find mislabeled data? We planted 711 wrong labels to measure it
We planted 711 wrong labels in 14,214 products. Cross-checking flagged 710; 95.4% of flags at confidence 0.9+ were planted errors. See the method.
Read the article ↗Can AI file vehicle complaints where NHTSA filed them? We measured it on 5,000 complaints from 2025 onward
StayCharted classified 5,000 later NHTSA complaints: 87.7% system and 85.9% subsystem accuracy. Compare models, confidence, mistakes and the test method.
Read the article ↗Part numbers from a photo: 100 look-alike parts, measured
StayCharted identified 87.8% of 1,000 unseen photos across 100 part numbers; 99.7% accuracy on the 32.3% above confidence 0.5. Compare both image models.
Read the article ↗How accurate is AI HS code classification? We measured it on 6,788 CBP rulings
We trained two models on 5,621 CBP rulings and tested them on 1,167 later ones. Accuracy at 10, 6, 4 and 2 digits, and how often the surest answers are right.
Read the article ↗Car photos for claims, service and OEMs: damage, damage type and part, measured
Compare StayCharted’s image models on car damage, 21 exterior parts and eight damage types. See measured accuracy, rare-category errors and review projections.
Read the article ↗Search relevance, measured: a model trained on past judgments against a general AI
Compare StayCharted and Claude on WANDS and ESCI: 4,000 query-product pairs per test, accuracy by category, review projections and train-test leakage.
Read the article ↗Can a model learn which access requests to approve?
StayCharted trained on 26,103 access decisions and tested on 6,666 held-out requests. Compare routine workload, review routing and missed denials.
Read the article ↗Can a model learn who should have access?
Compare trained models for access permissions. See held-out accuracy, review thresholds and results for learning group membership from past decisions.
Read the article ↗Asking an assistant vs. training a classifier: three benchmarks
Compare Claude Sonnet 5.5 with StayCharted trained classifiers on banking messages, contract clauses and medical abstracts. See all results, method and limits.
Read the article ↗Sorting medical research abstracts by condition: four approaches, measured
Medical abstract classification for life sciences literature triage: word models, embeddings and a fine-tuned model compared on published abstracts.
Read the article ↗Grading food by photo: fruit quality and cashew defects, measured
Food grading by photo with StayCharted AMT: compare fruit quality accuracy, cashew defect detection, and projected human review on FruitNet and VisA.
Read the article ↗Four ways to label contract clauses, measured on LEDGAR
Compare contract clause classification on LEDGAR: words, embeddings and fine-tuning across accuracy, training time and projected human review.
Read the article ↗PII Masking for AI Training: How AMT Checks Sensitive Data
Learn how personal information checks work, how masking applies to new text, and what pattern-based checks can miss.
Read the article ↗Image Classifier Benchmark: SigLIP 2 vs. DINOv2
SigLIP 2 scored 91.1% on catalog photos versus DINOv2’s 88.3–88.8%; DINOv2 base led on pet breeds. Compare StayCharted’s image-classifier tests.
Read the article ↗Why the AI Classifier reads words and meaning
See how AI Classifier Models compare meaning alone with words plus meaning and keep the better validation result. Explore measured results and tradeoffs.
Read the article ↗Banking Intent Classification: 93.7% Accuracy and Human Review
Banking Intent Classification with StayCharted AMT: 93.7% accuracy on the BANKING77 test set, plus confidence thresholds and human-review tradeoffs.
Read the article ↗STAYCHARTED AI MODEL TRAINER
Train AI to categorize the way your team does.
Start with examples you already have, see how often it’s right on items it never saw, and use confidence to prioritize the ones a person should review.
Start free