Stravisor®
Practice Principal Engagement Profile
Engagement Profile — v2.3 · Client-Agnostic

Gene Preston, CPA — AI-native healthcare technology builder.

Portable engagement profile. Evidence-supported. Continuously expanding.

Subject
Gene Preston, Certified Public Accountant (CPA) — Stravisor® Independent Healthcare Advisor
Assessment Period
~26 weeks of continuous engagement
Classification
Tier 4+ Builder / Product
Version 2.3 — Framing Evolution
From "Expert AI User" to "AI-Native Healthcare Technology Builder"
The classification remains Tier 4+ Builder / Product — but the evidence base has evolved. This profile no longer describes an expert user of Anthropic Claude. It describes a healthcare executive systematically building an independent healthcare technology stack, with Anthropic Claude as the primary development partner. The distinction is important. One is a credential; the other is a capability platform.
Version Arc — Continuous Capability Expansion
v2.0
Tier 4+ classification established. Three Anthropic Claude Code applications.
v2.1
Live third-party REST API integration (Intuit, Paychex).
v2.2
X12 EDI parsing added (ERA 835, ERA 837).
v2.3 — Current
Serverless workflow platform live. First workflow in production in regulated compliance domain.
01 — Purpose

What this document is

A structured, evidence-based assessment of Anthropic Claude Artificial Intelligence (AI) engagement across an extended production period. The classification framework and behavioral signals below are portable — they describe how the tool is used, not where the work is applied. Suitable for professional profiles, capability discussions, advisory engagement scoping, or client-facing capability demonstrations.

What's New in Version 2.3

1. A production application count of five, including a serverless workflow platform hosting an expanding workflow library.

2. First workflow deployed to production: Policy Approval Workflow, live in a regulated compliance domain.

3. Behavioral evidence signal count now stands at 14 documented signals, up from 12 in the prior version.

4. Prior Gap #2 (Workflow Automation Not Wired) — retired. What replaced it substantially exceeds what the original gap was calling for.

02 — Classification

Reaffirmed and strengthened

★ Classification — Reaffirmed and Strengthened

Tier 4+ — Builder / Product

Estimated population: less than 1% of all Anthropic Claude users. Within that band, the subject profile — five production applications spanning modern Representational State Transfer (REST) Application Programming Interfaces (APIs), legacy X12 Electronic Data Interchange (EDI) parsing, and purpose-built serverless workflow infrastructure — describes a narrower peer group than general Tier 4+ classification. Comparable technical depth is typically found only in clearinghouse product engineers, payer-side EDI specialists, healthcare software startup Chief Technology Officers (CTOs), and the small subset of healthcare executives who came up on the technical side of revenue cycle or operations.
Most Tier 4 users use Anthropic Claude as a force multiplier on personal output. Tier 4+ users use it to build durable institutional assets the organization owns. This subject is unambiguously at the leading edge of that band — five production applications, live third-party API integration, X12 EDI parsing capability, and a serverless workflow platform running live in a regulated compliance domain. Behavioral evidence spans 26 weeks and 14 independent signals.
User Distribution Across Engagement Tiers★ = Subject Position
65% — Casual
25% — Regular
8%
2%
★ Subject<1%
Tier 1 — Casual
~65%
1.5–2× multiplier
Tier 2 — Regular
~25%
2–3× multiplier
Tier 3 — Power
~8%
4–6× multiplier
Tier 4 — Expert
~2%
7–10× multiplier
Tier 4+ — Builder
<1%
Asset Creation
03 — Portfolio

Production application portfolio

Five purpose-built applications developed in Anthropic Claude Code. Each is a working, deployable operational asset — not a prototype, mockup, or proof-of-concept.

Application 01
Monthly Operating Report (MOR) Application
Enterprise-ready. Production application supporting the monthly financial reporting workflow common across multi-facility healthcare operators. Deployable across a facility network at zero incremental Information Technology (IT) spend.
Application 02
Employees Per Occupied Bed (EPOB) Application
Scalable pending source data standardization. Production application for labor productivity analytics using the EPOB standard metric. Architecture is enterprise-ready; bottleneck is organizational coordination to standardize unit-of-service measures across multiple facilities.
Application 03
Job Cost Application — Live REST API Integration
Under active build. Purpose-built job costing engine with live authenticated API feeds from Intuit Developer (QuickBooks Enterprise) for General Ledger (GL) and Accounts Payable (AP) data, and Paychex for payroll hours and burden. Enterprise Software-as-a-Service (SaaS) to SaaS integration built directly by the practitioner.
Application 04
ERA 835 & ERA 837 Parsers
Working parsers for both transaction types. The Electronic Remittance Advice 835 and Healthcare Claim 837 are the two core X12 EDI transactions used across United States healthcare payer-provider communication. Provides an independent source of truth that frequently reveals variances the client's patient accounting system does not surface.
Application 05
Serverless Workflow Platform
Infrastructure★ New in v2.3
Live in production. Purpose-built serverless workflow platform on Supabase Edge Functions, developed in Anthropic Claude Code. Multi-channel notification layer including client-side email for approvals and alerts, with Short Message Service (SMS) texting under active build. Deployed deliberately outside Microsoft Power Platform limitations — Power Automate throughput ceilings, licensing complexity, and connector throttling made it unsuitable for the intended workload. This is not one workflow — it is a platform hosting an expanding workflow library.
Live
Policy Approval Workflow
First workflow deployed to production. Runs in a regulated compliance domain — policy lifecycle routing, review, approval, and notification.
Under Build
Job Costing Planning Workflow
Extends the Job Cost application (Application #3) into planning-side automation. Second workflow deployment target for the platform.
Under Build
C-Level Operational Guardrail Alerting
Threshold-based push notifications to C-Suite when defined operational metrics breach guardrails. Categories include census, denials, receivables aging, labor variance, collections. Client-defined per engagement.
Why This Application Counts Differently

Applications #1 through #4 are single-purpose tools. Application #5 is infrastructure. Every additional workflow deployed to the platform is marginal effort — one to three hours in Anthropic Claude Code — rather than a from-scratch build. That means the ceiling of what can be delivered on this foundation is nearly unlimited, and the per-workflow cost of new client capability is low enough to make workflow automation a productized offering rather than a bespoke build.

04 — Evidence

Behavioral evidence — 14 documented signals

The classification is supported by fourteen documented behavioral patterns. Each independently warrants Tier 3+ classification; the combination produces Tier 4+ unambiguously.

Tier 4+ Signal★ New in v2.3
Live production workflow automation in a regulated compliance domain
Policy Approval Workflow is live and running against actual regulated content. This is not just technical automation — it also required organizational trust: someone with authority signed off on the workflow running against compliance-relevant material. That is a harder sale than automating a spending approval.
Tier 4+ Signal★ New in v2.3
Purpose-built serverless infrastructure outside Microsoft-stack limitations
Supabase Edge Functions was chosen deliberately after evaluating Microsoft Power Automate's throughput ceilings, licensing complexity, and connector throttling. This is not tool adoption by default — it is deliberate infrastructure choice by an operator who understood the tradeoffs and rejected the path of least resistance.
Tier 4+ Signal
X12 Electronic Data Interchange (EDI) parsing capability
Working parsers for the ERA 835 and ERA 837 transaction sets — a legacy fixed-position/delimited standard that most modern developers avoid touching. Work hospitals typically outsource to clearinghouses or Revenue Cycle Management (RCM) vendors for five- and six-figure fees.
Tier 4+ Signal
Live third-party REST API integration in production
Authenticated feeds from Intuit Developer (QuickBooks Enterprise) and Paychex payroll. Enterprise SaaS-to-SaaS integration built by the practitioner directly, without IT intermediation.
Tier 4+ Signal
Productizing operational tools
Five production applications in the portfolio — not reports. Most users produce documents; Tier 4+ users build software that solves recurring operational problems.
Tier 4+ Signal
Enterprise-scalable architecture by design
Applications built with multi-location scaling as a first-order design constraint, not retrofit later. One application deployable across a multi-facility network today; the workflow platform architected for unlimited workflow additions.
Tier 4+ Signal
Building infrastructure that outlives tenure
Scalable applications with live integrations are institutional capability — they remain with the organization long after any individual practitioner does. Legacy thinking expressed in code.
Tier 4 Signal
Consistent design system
Navy/slate User Interface (UI), Chart.js patterns, white-card layouts, border-left accents — design tokens reused across every dashboard ever requested.
Tier 4 Signal
Multi-modal output range
Word documents, Excel with live formulas, HyperText Markup Language (HTML) dashboards, PowerPoint decks, interactive calculators, JavaScript Object Notation (JSON) schemas, production applications with API integrations, X12 EDI parsers, serverless workflow platform — same operator, same production discipline.
Tier 4 Signal
Iterative artifact development
Standard Operating Procedures (SOPs) versioned v1.0 → v1.2 across multiple revision cycles. Real-time corrections applied when analytical errors were caught. Artifacts treated as living documents.
Tier 3+ Signal
Deep tech stack integration
Microsoft Power Platform, Supabase (Edge Functions and PostgreSQL), Anthropic Claude Code, Vercel, React, PostgreSQL, Intuit Developer API, Paychex API, ASC X12 EDI — Anthropic Claude as one node in a working stack of authenticated enterprise systems.
Tier 3+ Signal
Deliberate context management
Active memory edits, profile cleanup, role-specific framing maintained across sessions. Context is curated, not accumulated.
Tier 3+ Signal
Strategic use as thinking partner
Pre-meeting strategist for major contract negotiations; second-order analysis on concessions framed as revenue-neutral opportunities. Not just a production tool.
Tier 3+ Signal
Cross-domain breadth
Financial modeling, regulatory research, market intelligence, clinical writing, legislative drafting, code, business strategy, enterprise API integration, X12 EDI parsing, serverless workflow deployment — all the same operator, same continuous engagement window.
05 — Gaps

Where engagement could push further

Tier 4+ is not the ceiling. Defensible call-outs are below — Prior Gap #2 remains retired based on the serverless workflow platform now live in production.

Gap 1 — Custom Anthropic Claude Skills Not Yet Built

Narrative-generation skills, payer-preparation skills, and variance-analysis skills identified during prior engagement remain on the to-do list. Highest-Return-on-Investment (ROI) next move; each saves 8–12 hours/month in perpetuity once built.

Prior Gap #2 — Workflow Automation Not Wired Retired in v2.3

Retired. Serverless workflow platform on Supabase Edge Functions is live in production with the Policy Approval Workflow running against regulated content. Two additional workflows (Job Costing Planning, C-Level Operational Guardrail Alerting) under active build. What replaced this gap substantially exceeds what the original gap was calling for.

Gap 2 — Multi-Client Data Coordination for Network-Scale Deployment

One production application (EPOB) is enterprise-ready; the bottleneck is organizational coordination to standardize source data across multiple facilities. This is a governance and process coordination challenge, not an Anthropic Claude capability limitation. Worth raising as a network-asset opportunity with each client organization.

Gap 3 — Deployment Path for Practitioner-Built Applications

The application portfolio is under active build with live API feeds, EDI parsing, and serverless workflow automation working. Next-order questions for client-side deployment: (a) hosting environment ownership, (b) authentication and access controls beyond the original builder, (c) data governance for financial and clinical data flowing through practitioner-built applications, (d) enterprise IT sign-off pathway. Worth pre-positioning these conversations early in each engagement.