In partnership with

Automated Billing Eligibility Engine

ABI's engine integrates with PointClickCare to read chart data directly, evaluating each patient against a facility's custom billing eligibility profiles as a deterministic rules flowchart. Additional EHR integrations are in development.

How it's built

A rules engine that reads directly from EHR data

ABI connects to the EHR in read-only mode, runs chart data through a configurable rule flowchart, and logs every decision with its inputs. An optional AI layer can advise on documentation sufficiency and denial risk — it never makes the determination.

Data layer 01 / 03

Read-only ingestion & normalization

Scheduled sync from your EHR, designed to plug into any FHIR/REST source — live on PointClickCare today. Raw payloads are archived immutably so every decision can be reconstructed from the exact data it saw.

SourceEHR · read-onlyno write-back
SyncWebhooks + incremental pulls
StoreTenant-isolated · encrypted at restimmutable
Decision engine 02 / 03

Your rulebook as a flowchart

Every chart is walked through your billing playbook as a configurable flowchart. Deterministic nodes handle the clear-cut checks. An optional AI advisory layer flags documentation gaps and denial risk — logged alongside each determination, never replacing the human decision.

RulesDeterministic · versioned · testablecustom
AIOptional advisory layer, logged
ChangeRules evolve as your policy doeslive
Biller workflow 03 / 03

Review queue & decision log

Billers work from a queue organized by outcome — eligible, needs review, ineligible. Each item shows the rule path taken and the inputs evaluated. Every override is logged, and every decision is replayable against the exact data that produced it.

QueueEligible · review · ineligiblehuman-in-loop
TrailImmutable journal of every decision
ReportingDecision outcomes, overrides, denial patterns
Implementation

How a workflow gets built

01

Document the rules

We capture the existing billing rules — written policy, payer requirements, and institutional knowledge — and organize them into a structured rulebook.

02

Encode & connect

We translate the rulebook into a configurable flowchart, stand up a read-only EHR connection, and validate the logic against historical charts.

03

Run in parallel

The engine runs alongside existing reviewers. Each chart produces a decision and a full trace, compared against reviewer outcomes to verify accuracy.

04

Extend to new workflows

New service lines and payer types plug into the same infrastructure. The EHR connection and data layer stay shared — only the rules change.

Decision Path 7 Nodes
Start Evaluation
start
? Coverage Active Pass
coverage_active is_true
Verify Insurance Eligibility
call_insurance_eligibility
? Documentation Current Pass
docs_current is_true
Scan PCC Documents
scan_documents
? Meets Profile Criteria Pass
meets_criteria is_true
Decision
eligible
Evaluation Log Auto-Assessed
start Evaluation started for patient profile
check_coverage Condition 'Coverage Active?': True
Pass
call_insurance_eligibility Insurance eligibility verified with payer
check_documentation Condition 'Documentation Current?': True
Pass
scan_documents Scanned PCC documents for supporting evidence
check_criteria Condition 'Meets Profile Criteria?': True
Pass
eligible Decision reached: eligible
eligible
Recommendation: All profile criteria met — eligible
Analytics Pipeline reports
Time-to-Bill Funnel

Median and P90 durations for each pipeline stage, from chart sync through final determination, broken down per facility.

View report →
Eligibility Trends

Eligibility outcomes by profile, facility, and review cycle. Spot drift in determination rates before it affects collections.

View report →
Model Accuracy

Agreement rate between the deterministic flowchart and the advisory AI layer. Surfaces nodes where the two disagree most often.

View report →
Architecture

How the system is built

ABI is built on four properties that carry across every workflow it runs: consistent rule execution, a complete decision log, HIPAA-aware data handling, and infrastructure that scales without proportional reviewer headcount.

Consistency

Every chart runs through the same rule flowchart. The same inputs produce the same output, regardless of who is reviewing or when.

Auditability

Every decision is stored with its inputs, the rule path it took, and any AI reasoning. The full decision can be reconstructed at any point from the logged data.

Data handling

Encrypted in transit and at rest. Tenant-scoped access and facility-isolated storage. Record changes are archived under HIPAA versioning, never deleted. The system operates independently of the EHR and fails safely if connectivity is lost.

Volume independence

The rule engine processes charts the same way at any volume. Adding facilities or increasing case load doesn't require proportionally more review staff.

In production

Custom patient-profile eligibility running on PointClickCare

ABI is live in skilled nursing facilities through the PointClickCare Marketplace. Each facility's custom patient profiles — the criteria that define who qualifies, what evidence is required, and what timing windows apply — are encoded as a flowchart that runs against read-only FHIR and REST data from PCC.

Workflow

Per-resident eligibility review against facility-defined patient profiles

Each review cycle evaluates coverage status, census data, profile-specific criteria, reassessment windows, and documentation sufficiency for every resident. The engine encodes those steps as a flowchart connected to PCC's FHIR and REST feeds, producing a decision and an evidence set for every case.

PCC Marketplace Read-only FHIR + REST Multi-tenant SNF Custom patient profiles Auditable decision log
ConfigurableReassessment windows tracked automatically
Read-onlyGET-only PCC API · OAuth2 + mTLS
Per-profileCustom criteria per patient cohort
ImmutableEvery decision, override, and input logged
Biller review without automation

Constraints of Manual Billing Eligibility Reviews

Billing eligibility reviews are largely manual. Reviewers apply eligibility criteria to chart data, assess compliance windows, and document decisions — a process that varies by reviewer, organization, and service line.

01

Manual review

Eligibility determinations are made by individual reviewers working through chart data, payer rules, and documentation requirements case by case.

02

Compliance windows

Documentation, assessments, and orders each carry their own deadlines. Missing any one of them can invalidate an otherwise eligible claim.

03

Audit reconstruction

When a claim is denied, reconstructing the original eligibility determination often requires reviewing chart notes and applying judgment after the fact.

04

Review capacity

The volume of reviews a billing team can complete is tied directly to headcount. Adding facilities typically means adding reviewers.

ABI is live on PointClickCare, automating eligibility review against each facility's custom patient profiles. Additional EHR integrations are in development. If you're working on an eligibility automation problem in healthcare, we're open to a conversation.

FAQs

Frequently Asked Questions

If yours isn't listed, please feel free to reach out directly.

No. ABI is strictly read-only. We ingest what's already there, make decisions outside your system of record, and never modify records, orders, or treatment plans. If ABI is down, your EHR and facility operations continue normally.
No. ABI is an eligibility automation engine — it produces the determination and the evidence that supports it. Your billing team or clearinghouse handles the submission.
PointClickCare is live (FHIR + REST, OAuth2 with mTLS). The architecture treats the EHR as a pluggable source — additional systems can be added on request.
Core eligibility is deterministic — the flowchart runs the same way every time. An optional AI advisory layer (Gemini 2.0 Flash via Vertex AI) can evaluate documentation sufficiency and flag denial risk, with reasoning logged for every call. The AI never makes the final determination.
Onboarding runs in five stages: playbook capture, rulebook encoding, read-only data connection, parallel run against historical charts, then live biller queue. Total timeline depends on the complexity of your playbook and the number of payers involved.
Encrypted in transit and at rest. Tenant-scoped access and facility-isolated storage. Record changes are archived under HIPAA versioning, never deleted. Sync-log retention is aligned to compliance requirements.