8 items need your attention today. Everything else is being handled automatically.
How the system handles 26,000 providers without requiring proportional human effort.
Total credentials monitored
68,400
Resolved automatically (last 30d)
1,847
Required human intervention
127
Automated resolution rate by credential type
Why cases reach the human queue
Post-expiration decision
38%
4+ messages, no response
29%
Payor credentialing renewal
21%
Document anomaly flagged
12%
AI outreach sequence — what happens automatically
Day 0 — 90 days before expiration
AI sends heads-up message AI
Personalized to provider, credential type, state. No human approval needed.
Provider clicks renewal link
Provider submits structured renewal form provider
Enters new expiration date, uploads certificate. Form linked directly to database.
Immediately on submission
AI verifies submission automatically AI
Checks expiration logic, document present, license format. Pass → database updates. Fail → human queue.
Credential verified
Database updated, flag cleared, case closed fully automated
Zero human involvement. ~88% of state license and insurance renewals.
If no response at 75 days
AI sends second message AI
Automatically triggered. No human action needed.
If no response at 60 days
AI sends third message, critical tone AI
Final automated attempt before human escalation.
If no response at 45 days
Enters human task queue human required
Full history assembled. Ops person reaches out directly.
If credential expires
Immediate escalation — session pause decision human required
Only a human makes this call.
Six-month view — January through June 2026.
Automated resolution rate
88%
▲ +18pts since Jan
Avg days to resolution
11.2d
▼ from 34d in Jan
Human hours saved / mo
~340 hrs
▲ growing with network
Anomalies caught by AI
47
Would have slipped through manually
Automated resolution rate
% resolved with zero human involvement
Avg days to resolution
From first AI outreach to credential verified
Human queue vs credentials processed
Queue stays flat as network grows
Human hours saved per month
Based on 12 min avg handling time per credential
What the trends are telling us
Resolution rate plateauing at ~88–89%
Expected ceiling — payor credentialing (43% auto) pulls average down.
Days to resolution: 34 → 11
Providers learning to use the renewal form month over month.
Human queue flat at 6–9 items/day
Network grew 12% in 6 months. Queue did not grow. Scalability confirmed.
47 document anomalies caught by AI
All would have passed manual review. Includes 3 potentially fraudulent submissions.
Hours saved: 89/mo → 340/mo
At $45/hr ops cost: ~$15,300/mo saved in June alone.