aj Healthcare Infrastructure Research — Executive Summary

Healthcare Infrastructure API Research

Prepared for
Arthur J. Gallagher Co.
February 2026

Executive summary of research into healthcare data intermediaries — the infrastructure layer that could power personalised benefit recommendations for Gallagher’s mid-market digital engagement solution.

01

Why We’re Looking At This

Gallagher serves employers across hundreds of health plans, dental carriers, vision networks, and ancillary benefit programs. Today, connecting to each carrier individually would require building and maintaining hundreds of point-to-point integrations — a cost and complexity model that doesn’t scale. It was identified that what we need is an intermediary layer: a set of APIs that sit between Gallagher and the carrier ecosystem, normalising data so we can deliver a single, coherent benefit experience to members. Think of it as the Plaid equivalent for health benefits — a connectivity layer that abstracts away the chaos underneath.

02

What We Found

The market has split into two distinct camps. On one side: API-native companies like Flexpa, Ideon, and Noyo that were built from the ground up as developer-friendly data rails — clean REST APIs, FHIR-based data models, modern OAuth consent flows. On the other: legacy network operators like Optum (Change Healthcare) that process enormous transaction volumes through decades-old EDI pipelines, now layering APIs on top.

The Plaid-equivalent layer hypothesised actually exists — but it’s not a single product from a single vendor. The capability is distributed across multiple specialists, each owning a different slice of the data stack. No one vendor covers claims and plan design and enrollment execution. The architecture that emerges is composable: pick the best rail for each layer and connect them through a proprietary orchestration layer that Gallagher owns.

Key Insight
The proprietary value — the recommendation engine that tells a member “switch to the plan with glasses coverage to save $200/year” — should be built and owned by Gallagher. The vendors below provide the data plumbing that feeds that engine. This is where our differentiation lives.
03

Recommended Stack

Four layers, each with a lead vendor and a clear responsibility boundary. The layers are intentionally independent — if any single vendor underperforms or gets acquired, the others remain unaffected.

Layer Vendor Role
Layer A
Flexpa
Member data ingestion & consent. Connects to payer systems via FHIR Patient Access APIs with member-directed OAuth consent. Retrieves claims history, EOBs, coverage details, and provider data. This is the pipe that answers “what has this member actually used and spent?”
Layer B
Ideon
Plan design intelligence. Machine-readable plan data across medical, dental, and vision — deductibles, copays, coinsurance, formulary tiers, network configurations. Powers the comparison logic: “here’s what Plan A covers vs Plan B for your usage pattern.”
Layer C
Noyo
Enrollment execution. When a member accepts a recommendation, Noyo handles the enrollment change — translating API calls into carrier-accepted 834 transactions. Covers adds, terms, plan changes, and dependent updates with real-time confirmation.
Layer D
Arcadia
Population analytics & longitudinal modelling. Aggregates data across the member population for trend analysis, risk stratification, and employer-level reporting. Helps answer “across all 5,000 members in this group, where are the biggest savings opportunities?”
The Gallagher-Built Layer
Sitting above these four rails: the recommendation engine and member experience that Gallagher builds and owns. This is the orchestration logic, the savings calculations, the plan comparison UI, and the personalised nudges. It consumes data from Layers A–D and produces the “360 benefit view” that differentiates our offering. This intellectual property stays with us.
04

What This Means For Gallagher

This architecture gives us a path to the mid-market digital engagement solution without building carrier connectivity from scratch. Instead of spending 18–24 months and significant engineering resources on payer integrations, we plug into existing rails and focus our build effort on the recommendation logic and member experience — the parts that create defensible value.

Concretely, this means we could reach a first proof-of-concept with a single employer group in roughly 3–4 months: Flexpa for claims data, Ideon for plan design, and a lightweight recommendation engine on top. Enrollment execution (Noyo) and population analytics (Arcadia) layer in during Phase 2 once the core recommendation loop is proven.

The composable approach also protects us strategically. We’re not locked into any single vendor. If Flexpa gets acquired or Ideon changes pricing, we swap that layer without rebuilding the whole stack. This is the same principle behind the API-first architectures that have worked well in fintech — own the experience layer, rent the plumbing.

05

What We Don’t Know Yet

This research is based on public documentation, developer docs, regulatory filings, and published case studies. Several material questions remain open and will require direct vendor engagement to resolve.

  • $
    Pricing at Gallagher scale. Most vendors don’t publish PMPM rates. Flexpa, Ideon, and Noyo all require sales conversations for enterprise pricing. We need actual quotes based on our member volume to validate the $2.4M–$4.2M Year 1 estimate in the detailed reports.
  • Production onboarding timelines. Sandbox access and first-data timelines are documented, but production deployment with real carrier connections is less clear. Flexpa’s FHIR connections depend on payer implementation quality, which varies.
  • Broker-of-record data access. Can Gallagher’s broker status unlock employer-level data feeds without individual member consent for each connection? This affects whether the data flow is push (employer-authorised) or pull (member-directed), which has major UX and adoption implications.
  • Multi-line aggregation. No single vendor has demonstrated medical + dental + vision + pharmacy data flowing through one normalised pipeline in production. The composable stack assumes this is solvable, but it hasn’t been proven at our scale.
  • Optum’s strategic posture. Optum (Change Healthcare) has the deepest payer connectivity but carries significant baggage — DOJ antitrust scrutiny, the 192.7M record breach, and structural conflicts as a UHC subsidiary serving all-carrier clients. Whether to include them at all requires a risk-appetite conversation.
06

Recommended Next Steps

Three concrete actions to move from research to validation. Each is designed to de-risk the architecture before committing budget.

1 Pilot vendor conversations: Flexpa & Ideon
Request sandbox access and enterprise pricing from both. The specific ask: “Given a 50,000-member employer group with Aetna/BCBS/UHC mix, show us (a) what claims data Flexpa returns via FHIR, (b) what plan design data Ideon provides for those carriers, and (c) PMPM pricing at that volume.” This validates whether the two core layers deliver recommendation-grade data.
2 Internal alignment on data consent model
Before engaging vendors, resolve whether the data flow is member-directed (each person authorises their own data pull) or employer-authorised (broker-of-record status enables bulk access). This is a legal and product decision that shapes which vendor capabilities matter most. Suggest a working session with compliance, legal, and product.
3 Proof-of-concept scope definition
Define a minimal first use case — e.g., “for members in Group X, show whether they’d save money switching dental plans based on their claims history.” Constrained scope (one employer, one benefit line, one recommendation type) lets us validate the full data pipeline end-to-end without boiling the ocean.
07

Full Reports

The research behind this summary lives in four detailed reports. Round 1 maps the full landscape; Round 2 pressure-tests the shortlisted vendors against our specific use case.

Additionally, the initial benefits vendor landscape survey (covering wellness, mental health, and financial wellbeing platforms) is available as a separate reference:

Reference
Health & Wellness Vendor Landscape
Claude
Broader benefits ecosystem survey covering wellness, mental health, financial wellbeing, and engagement platforms. Context for how infrastructure vendors fit into the full picture.
View report →