Confidential client, name withheld under NDA

Cut Prior Authorization Turnaround From 5 Days to 1.5 Days for a US Multi-specialty Healthcare Group

Quick answer: An AI agent can pull chart data from Epic through FHIR APIs, check payer rules, flag missing documents and draft payer-ready prior authorization requests. Bacancy Technology built this for a US multi-clinic group.

Outcomes deliveredDelivered February to April 2026 by a team of 3 forward-deployed engineers
1.5 day

Average prior auth turnaround

Down from 5 days

70%

Less manual prior auth work

Was 25 to 40 minutes per request

40%

Fewer documentation-related denials

Gaps flagged in Epic before submission

95%

Coordinator adoption in month 1

Live across all clinics in 12 weeks

Client name withheld under NDA. To honour our non-disclosure agreement and compliance obligations, we describe this client by industry and region only. The project scope, technology and outcomes are shared without details that could identify them.

About the Client

The client is a US multi-specialty healthcare group offering orthopedic, cardiology and pain management care across more than a dozen clinics and over 100,000 patients a year. Coordinators copied patient data from Epic into payer portals by hand, and missing documents led to frequent denials. A tool from a previous vendor was abandoned within 3 months.

Bacancy Technology embedded forward-deployed engineers with the client's coordinators, billing leads and compliance team. They built an AI agent integrated with Epic and rolled it out to every clinic in 12 weeks. Average turnaround fell from 5 days to 1.5 days.

Client type
Multi-specialty healthcare group
Industry
Healthcare
Location
USA
Team
3 forward-deployed engineers
Timeline
February 2026 to April 2026
Project size
Non-disclosable
Engagement
Forward-deployed engineering

Four problems stalling prior authorizations

Manual data entry, late documentation gaps and a failed first tool left the client with a 5-day prior authorization backlog.

Manual payer portal entry

What was happening

Coordinators spent 25 to 40 minutes per request copying codes, notes and imaging reports from Epic.

Business Impact: The growing backlog stretched average turnaround to 5 days.

Late documentation gaps

What was happening

Missing clinical documents caused a large share of denials, found only after the payer rejected them.

Business Impact: Resubmissions delayed care and forced rescheduled procedures.

Tool built off-site

What was happening

The previous AI team worked from a requirements document and never visited the clinics.

Business Impact: Payer rules in coordinator notes were missed and the tool fit no daily routine.

No owner for adoption

What was happening

The previous team handed over the tool and left, with manual Epic exports and no staff training.

Business Impact: Usage dropped to almost zero within 3 months.

Challenges & Solutions

Our forward-deployed engineers worked inside the clinics and owned each problem through to adoption.

Challenge 1: Capturing real clinic workflows and payer rules

The previous team missed payer rules buried in coordinator notes and built a tool that did not match clinic routines.

Solution: Our Forward Deployed Engineers spent 2 weeks with front-desk staff, coordinators and billing leads at 2 locations. They mapped every step from order to payer decision. They turned the rules for the top 12 payers into a rules layer that billing leads update on their own. It covers Medicare Advantage, Medicaid managed care and major commercial insurers.

Challenge 2: Removing manual data entry from Epic to payer portals

Coordinators spent 25 to 40 minutes per prior authorization copying diagnosis codes, notes and imaging reports by hand.

Solution: Our engineers built a FHIR API connection to Epic for the AI agent. It pulls ICD-10 and CPT codes, clinical notes, lab results and imaging reports in real time. The agent then drafts a payer-ready request with a clear medical necessity statement.

Challenge 3: Catching missing documentation before submission

Coordinators learned about missing clinical documents only after a payer rejected the request.

Solution: Our team added a check that tests each draft against payer criteria before submission. When a document is missing, the care team gets an alert inside Epic, so gaps are fixed before the payer sees the request.

Challenge 4: Earning compliance approval and lasting adoption

The previous team left with no compliance sign-off, no training and no one accountable for usage.

Solution: Our engineers worked with the client's compliance and IT teams from week 1, and the compliance team approved the system before rollout. The agent follows HIPAA's minimum necessary standard, runs on BAA-covered Azure hosting and logs every data access event. After a pilot at 1 orthopedic clinic, our team trained staff on-site at every location and held weekly reviews after go-live.

Agent workflow

How the prior authorization agent works

Each request moves from Epic chart data to a checked draft, and coordinators approve every request before it goes to the payer.

  1. The client's AI agent system
    1. Gather
      Pull chart dataEpic FHIR APIs, in real time
    2. Check
      Payer rules layerTop 12 payers
      Missing document checkAlert inside Epic
    3. Draft
      Draft requestMedical necessity statement
    4. Review
      Coordinator reviewApprove or edit the draft
    5. Track
      Status trackingAll pending authorizations

Every data access event is logged for compliance audit.

How the Platform Works

The client prior authorization dashboard

Core Features

  • Real-time Epic chart retrievalCodes, notes, labs and imaging pulled through Epic FHIR APIs.
  • AI-drafted prior auth requestsPayer-ready drafts with a medical necessity statement, ready for review.
  • Payer rules layerRules for the top 12 payers that billing leads update on their own.
  • Missing documentation alertsGaps flagged inside Epic before a request reaches the payer.
  • Coordinator review workspaceCo-designed with clinic staff for quick approval, edits and status tracking.

The stack behind the agent

Frontend
React | TypeScript
Backend
Python 3.11 | Node.js
AI and ML
Anthropic Claude
Integrations
Epic FHIR APIs | HL7 | Payer portal automation
Database
PostgreSQL | Redis
Cloud and DevOps
Microsoft Azure (HIPAA-eligible services) | Docker | Azure Kubernetes Service
Security
BAA-covered hosting | PHI encryption | Role-based access | Audit logging

Ready to start?

2500+ projects delivered by Bacancy Technology

Get access to an experienced team of developers and engineers from Bacancy Technology, handpicked to ace your goals. Kickstart within 48 hours, no-risk trial.

Book A 30 Min Call
14+Years of business experience
1458+Happy customers
30+Countries with happy customers
1050+Agile-enabled employees

More Case Studies

View All Case Studies