Human-in-the-Loop
Approvals & Supervision
Humans in the loop, exactly where it matters.
Automation is powerful, but trust requires oversight. Filing Cab’s approval system ensures that AI decisions are reviewed when confidence is low, stakes are high, or policy demands it. Every automated action can be supervised, and every human decision feeds back into the system to improve future accuracy.
Confidence-Based Routing
Set confidence thresholds per document type, per action, or globally. Documents that exceed the threshold are processed automatically. Those below it are routed to review queues with the AI’s best guess, alternative candidates, and the evidence that informed each option. Reviewers make informed decisions, not blind ones.
Review Queues & Workflows
Organize pending reviews by urgency, document type, source, or assigned reviewer. Route specific document types to specialist reviewers — legal documents to legal, financial documents to accounting. Set SLAs and escalation rules so nothing sits in a queue longer than it should. Batch review interfaces let reviewers process high volumes efficiently.
Full Audit Trails
Every action — automated or human — is logged with timestamps, actor attribution, and the decision context. Who classified this document? What confidence score did the AI assign? Who approved the override? When was it filed, and what rule triggered the routing? Complete traceability from ingestion to final disposition.
Escalation & Override Policies
Define escalation chains for documents that require multi-level approval. Sensitive filings can require manager sign-off. Unusual patterns — a sudden spike in a document type, filings from a new sender, documents that contradict existing records — can trigger automatic escalation to senior reviewers or compliance officers.