Services

Four ways to work together, from a short assessment to building and running software for you. Every engagement is scoped to a clear outcome.

AI readiness assessment

For agencies that know AI matters but don't know where to start, or have tried tools that didn't stick.

Usually 2 to 3 weeks.

What's included

  • Interviews with intake, clinical, scheduling and billing leads
  • Review of your EMR, data sources and current reporting
  • Security and PHI review of any planned AI use
  • Ranked use cases with value and effort estimates

You walk away with

  • A 90-day roadmap and a ready-to-build pilot plan

AI agent pilots and builds

For a specific workflow you want automated: eligibility, intake, EVV exceptions, denials, outbound calls.

Usually 6 to 10 weeks per pilot.

What's included

  • Agent design in Python, Azure AI Foundry or your preferred platform
  • Integration with your EMR, clearinghouse or payer portals
  • Approval steps, logging and monitoring
  • Baseline and results measurement

You walk away with

  • A working agent in production, with documentation and staff training

Full-stack software

For portals, internal tools and integrations your off-the-shelf systems don't cover.

Scoped per project.

What's included

  • Angular front ends with .NET / C# APIs, or Python services
  • Azure SQL, Entity Framework and clean, testable architecture
  • Azure Functions, Container Apps and Data Factory pipelines
  • CI/CD with Azure DevOps or GitHub

You walk away with

  • Source code you own, deployed in your environment

Analytics and data platforms

For leaders who want one trusted view of census, visits, staffing and revenue.

Scoped per project, or ongoing.

What's included

  • Microsoft Fabric or Azure SQL data platforms
  • Power BI semantic models with agreed KPI definitions
  • Dashboards for operations, clinical quality and revenue cycle
  • Data quality checks and refresh monitoring

You walk away with

  • Reports your teams use every week, and the skills to maintain them

Fractional AI and data lead

For agencies that need senior technical leadership without a full-time hire.

Monthly retainer.

What's included

  • Vendor and tool evaluations, including AI products pitched to you
  • Roadmap ownership and steering with your leadership
  • Oversight of internal developers or contractors
  • Board- and owner-ready updates

How I deliver

The frameworks behind every engagement, from the data platform to the way a single request becomes working software.

From scattered data to decisions

Every source lands in one governed pipeline, and every output reads from it.

Sources

  • EHR and EMR data
  • Data warehouse and operational data
  • Snowflake and cloud warehouses
  • Spreadsheets and generated reports
  • Manual entry
  • Web and portal data
  • Payer and partner APIs

Automated ingestion and cleansing

  • Structured data
  • Unstructured documents
  • Validation, matching and deduplication
  • Curated, governed datasets

Outputs

  • Presentation and reporting
  • Executive insights in Power BI
  • Apps and data services
  • AI reasoning with LLMs

Value delivered

  • Real-time visibility
  • Actionable insights
  • Strategic decision-making
  • Operational efficiency

Reference solution architecture

A pattern I've built on Azure: web apps and services around one data core, with AI agents and external data at the edges.

Data augmentation

  • Partner and payer APIs
  • AI agents (Python)

Azure Functions

  • Triggers
  • Integrations

Web apps (Angular)

  • Sales
  • Intake
  • Provider management
  • Payer management
  • Revenue cycle

Power Platform

  • Datasets
  • RPA
  • Pipelines
  • Dashboards

APIs (.NET)

  • Referrals
  • Clients and patients
  • Documents

Infrastructure

  • Security and identity
  • Logging and error handling

AI services

  • MCP servers
  • Microsoft Foundry

Azure SQL

  • Inbound and staging
  • Patients, clients and events
  • Analytics

Storage

  • Documents
  • Data extracts
  • Processing artifacts

External data

  • SQL sources
  • Files and CSV
  • Excel
  • APIs

AI technology

  • Microsoft Foundry
  • Claude
  • Gemini
  • Local LLMs (Ollama)
  • LangChain
  • n8n

Core technology

  • .NET Core
  • Angular
  • Python
  • Azure SQL
  • Azure Storage
  • Azure Functions
  • Event Grid
  • Power Platform

Data maturity assessment

An example from an assessment. Many organizations have built more capability than the business uses. The gap between the two dots is value already paid for.

1 Initial2 Repeatable3 Defined4 Measured5 OptimizedArchitecture and integrationOperational data store, pipelines, lakehouseData quality and master dataClient, payer and episode recordsGovernance and definitionsLineage, access, PHI handlingReporting and self-servicePower BI and semantic modelsAdvanced analytics and AIPrediction and automationOperating modelOwnership, intake, practice

Platform capability Business realization Unrealized capability

Analytics delivery lifecycle

Two loops: data engineering builds curated datasets once, and report development reuses them for every request.

Data ingestion

  1. Information requirements
  2. Data sources
  3. Data profiling
  4. Source-to-target mapping
  5. Data quality assurance
  6. Dataset preparation and semantics
Curated datasets

Dashboard and report development

  1. Requirements and user stories: what insight do you need?
  2. Design: visuals, navigation and interaction
  3. QA: verify the measures are accurate
  4. UAT: users accept and validate
  5. Release: publish, train, support
Insight

Published results generate new requests, and feedback from UAT flows back to the data stage.

How one request becomes user stories

Every analytics request is checked against each layer of the platform, so nothing is missed and estimates are honest.

PresentationDashboards and reports
Subject areasSales, Intake, Clinical operations, Outcomes, Billing and finance
ProcessStage, cleanse, normalize, load
Star schemaFact with Date, Payer, Patient and Provider dimensions
SourcesEHR and EMR, structured and unstructured data
  1. Extract Do we need data from a new source?
  2. Load and transform Do we need to load or transform data?
  3. Data model Does this need a new model or a change to the existing one?
  4. KPI and calculation Does it need a new KPI, calculation or analytic?
  5. Dashboard Does it need a new dashboard or report?
  6. QA A story for testing the results.
  7. Deploy A story for productionizing and deploying.

Sprint cycle

Two-week sprints with a working increment at the end of each.

  1. Product backlogAnalytics and feature stories, ranked by value
  2. Sprint planningPick the stories for this sprint
  3. SprintDesign, implement, test, deliver, with a daily stand-up on progress, blockers and plans
  4. Review and retrospectiveDemo the increment, then improve how we work

Automation lifecycle

How a manual process becomes a reliable bot or agent.

  1. IdentifyFind high-volume, rule-based work
  2. AnalyzeMap the steps, exceptions and data
  3. DesignDefine the flow and the human checkpoints
  4. DevelopBuild the bot, agent or integration
  5. TestRun against real cases with staff
  6. ImplementDeploy, monitor, and hand over

Find the first AI win in your agency.

A free 30-minute call. Bring your messiest workflow.

Book a free consult