Skip to content

Case Study

Epic Onboarding at Wire Speed

Alex maps every new hire to the right Epic template overnight, so the access decision is made before the team starts its day.

A group of doctors with stethoscopes and iPads are gathering around to discuss an issue
20%

Daily time saved

Daily

Cadence, ready each morning

0

Integration projects required

3 min read

Key takeaways

  • The bottleneck was never writing access into Epic. It was deciding which template each new hire should receive.
  • Alex is deliberately not connected to Epic. He hands over a reviewable file, so no integration project was needed.
  • The health system gained 20% daily time savings, with the analysis ready before the team starts work each morning.
  • A human approves every grant, which is what keeps over-provisioning of protected health records out of the process.

About the Organization#

A large regional health system with more than 20 hospitals, over 400 clinics, delivering acute, pediatric, and behavioral care across multiple states, under strict federal and state health-data regulation.

In a health system, almost everyone needs Epic access before they can do their job, and getting it right matters in three directions at once. It is critical: a clinician waiting on access cannot see patients or chart. It is high-volume: every new hire needs it, continuously. And it must be precise: each person should see exactly what their role requires, and nothing more.

Epic makes access precise through a permission model of templates and sub templates. The hard part is deciding which template each new hire should get, based on department, site, the specific hospital, and role. At the health system, that decision was manual: one onboarding ticket at a time, done by hand.

Why This Matters in Healthcare#

Manual template mapping: For every ticket, an expert gathered a new hire’s attributes and searched the template register by hand to find the right match.

Throughput ceiling: The rules were clear and the expertise was there. The limit was how much judgment-heavy work one team can do by hand as new hires keep arriving.

Care gated by access: Delays in provisioning meant clinicians could be left unable to see patients or chart on day one.

Precision under regulation: Over-provisioning a single grant risks exposing protected health records. Every mapping has to be exactly right.

How Alex Automates Epic Provisioning#

The health system put Alex, Twine’s first AI Digital Employee, to work alongside its Access Management team. Alex learned how the organization maps attributes to access and now runs onboarding mapping as a scheduled daily agentic task, so the analysis is done before the team starts its day.

Notably, Alex is not connected to Epic and does not need to be: he does the decision work and prepares the exact file Epic can ingest, and an expert reviews and imports it. The value arrived without waiting on an integration project.

How It Works#

Each night, Alex reads the open onboarding tickets and pulls the four attributes that decide Epic access: department, site, hospital, and role. He matches that profile against the organization’s template register, applying the same mapping rules the Access Management team already follows by hand. The output is a single reviewable file of template and sub-template assignments, waiting for an expert to check and import when the team starts its day.

Alex reads each new hire’s department, site, hospital and role from the onboarding ticket, then matches them against the Epic template register.
The daily output: a reviewable file of template and sub-template assignments that an Access Management expert checks and imports into Epic.

An Organization Expert Stays in Control#

This is human-in-the-loop by design. Alex does the heavy, repetitive analysis; an organization expert makes the final call. Because Alex hands off a file rather than writing to Epic, the expert is the last check before any permission exists.

The mapping itself is transparent and tunable, driven by reference files and explicit instructions rather than a black box, so the organization can see and adjust how a template is chosen. And it all runs on top of the organization’s existing Epic setup, with no rip-and-replace.

Why Not a Native IGA Epic Connector?#

IGA platforms can provision templates directly into Epic, however a connector only writes access once something has decided the correct template and sub template for each person. That decision, mapped from department, site, hospital, and role, is the real bottleneck.

Standing up and maintaining that logic, plus the Epic web-services integration itself, is a significant workload, something the organization wanted to avoid.

Alex closed the gap first: he makes the mapping decision, hands the organization a reviewable file, and keeps a person in control of every grant, with no integration project required.

Questions We Get a Lot

Ready to Hand the Identity Backlog to Alex?

Alex, Twine's AI Digital Employee