Customer Support Bot v2

End-to-end evaluation for the customer support chatbot including policy compliance, accuracy, and faithfulness checks.

PASS

128

Public

run-001

OpenAI

Auth: Server KeyModel: gpt-4oHealth: Not Tested

Pass Rate Min90%

Faithfulness Min0.85

PII Max0

Secrets Max0

Jailbreak Max0

PIIBlocking

SecretsBlocking

JailbreakBlocking

ToxicityBlocking

Case Uploads

Status	File	Template	Cases	Date
SUCCESS	support_cases_batch_1.csv	Standard Eval Cases	45	Dec 5, 2025

Result	Run ID	Trigger	Model	Execution	Pass Rate	Accuracy Delta	Started
PASS	run-001	Nightly	gpt-4o	Simulated	95.3%	+1.2%	2d ago
PASS	run-002	PR	gpt-4o	Simulated	93.8%	-0.3%	3d ago
WARN	run-009	Manual	gpt-4o	Simulated	90.6%	-1.5%	6d ago