NeuralSeek · Governance & Orchestration Layer

R. Almánzar

Back to connectors

Fintech Lab · Fraud Scoring Model

https://lab-fraud.apse.bancocaribe.aws/score

healthy

Calls (24h)

149,600

p95 latency

88ms

Error rate

0.2%

Last sync

2m ago

API management

Routing and access, managed above the bank's own API baseline.

Layer
AI touchpoint
Protocol
gRPC
Authentication
IAM role assumption
Endpoint
https://lab-fraud.apse.bancocaribe.aws/score
Retry policy
3 attempts, exponential backoff, 8s ceiling
Rate limit
600 requests / minute per screening worker
Data handling
PII masked at capture; payloads retained for audit replay

Existing lab model, now called through the governance layer.

Where this connector is used

Recent screenings that called it — click through to the audit.

This connector fronts an AI touchpoint — Fraud Scoring Model · currently ungoverned