Medical records processing · Production

Notable Health AI Agents reduce faxed referral turnaround from 48 hours to 10 minutes at Florida health system

The problem

A large Florida health system relied on manual fax processing by 12.5 full-time employees, with an average turnaround of 48 hours that could reach five business days when staff were absent. Fifteen percent of referrals were handwritten, and faxes arrived across multiple queues making volume and throughput tracking difficult.

Workflow diagram · grounded in source
1
Faxed referral received
Trigger
Inbound faxed referrals arrive across three fax lines deployed in the health system's contact center.
▾ source quote
“deployed Notable's Referrals Coordinator AI Agent across three fax lines in its contact center”
2
AI data extraction
Ai action
Notable's AI Agent extracts necessary details from both structured and unstructured fax data, including handwritten clinician notes.
▾ source quote
“For structured data, it connects directly to databases or APIs within EHR systems to retrieve information from specific fields; for unstructured data, such as a faxed referral document or a clinician's handwritten note, the Agent can extract the necessary details …”
3
Order validation and flagging
Validation
The AI Agent flags duplicate orders, orders that fail medical necessity, or orders missing information from the referring provider.
▾ source quote
“It can also flag duplicate orders, orders that fail medical necessity, or orders that are missing information from the referring provider”
4
Complex case routing to humans
Routing
Only the most complex cases are routed to humans for review while straightforward orders are handled autonomously.
▾ source quote
“only routing the most complex cases to humans for review”
5
Order transcription completed
Output
Automated order transcription is completed for 14 different order types including MRI, x-ray, CT, ultrasound, and others.
▾ source quote
“automating the order entry of 14 different order types: MRI, x-ray, CT, ultrasound, mammography, DEXA, CTA, PFT, nuclear medicine, barium swallow, IR, lab, ambulatory, and PET”
Reported outcome

Notable's AI Agents automated over 10,000 faxed orders to date, saving 8,000 hours annually with a 2.6x ROI, reducing average turnaround from 48 hours to 10 minutes, achieving an 85% referral completion rate, and enabling 5 FTEs to upskill for higher-value work.

Reported metrics
Faxed orders automated to dateover 10,000
Projected total faxed orders automated in 202560,000
Staff hours saved annually8,000 hours annually
ROI on referral transcription2.6x
Show all 11 reported metrics
faxed orders automated to dateover 10,000
projected total faxed orders automated in 202560,000
staff hours saved annually8,000 hours annually
ROI on referral transcription2.6x
FTEs upskilled for higher-value work5
referral completion rate85%
turnaround time before automation48 hours
turnaround time after automation10 minutes
FTEs previously doing fax indexing12.5
handwritten faxed referrals proportion15%
worst-case turnaround when staff absentfive business days
Reported stack
Notable's Referrals Coordinator AI AgentNotable's AI PlatformRightFaxOnBaseEpic
◆ Does this fit your context?

Compare to your context

Tell us your scale, team, and constraints. We'll show what changes at your size, what fails at your scale, and whether this case is a fit, needs adaptation, or won't scale to you. Free demo, no signup.

Compare to your context →
~30 seconds · free
Source
https://www.notablehealth.com/customer-stories/optimizing-referral-management-with-ai
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Notable's AI Agents automated over 10,000 faxed orders to date, saving 8,000 hours annually with a 2.6x ROI, reducing average turnaround from 48 hours to 10 minutes, achieving an 85% referral completion rate, and enab…

What tools did this team use?

Notable's Referrals Coordinator AI Agent, Notable's AI Platform, RightFax, OnBase, Epic.

What results were reported?

Faxed orders automated to date: over 10,000; Projected total faxed orders automated in 2025: 60,000; Staff hours saved annually: 8,000 hours annually; ROI on referral transcription: 2.6x (source-reported, not independently verified).

How is this medical records processing AI workflow structured?

Faxed referral received → AI data extraction → Order validation and flagging → Complex case routing to humans → Order transcription completed.

Related medical records processing cases

WHAT TO DO WITH THIS

Now compare it to your context

This case is one data point. Whether its pattern fits you depends on your volumes, your stack, and your exception load — that comparison is the step no case study can do for you.

Browse the Medical records processing cases →