Client retained at $16K/month · Phase 3: Predictive no-show modeling + AI-assisted clinical documentation (in progress)
Who we worked with
A hospital network operating 4 facilities across the United States, managing 200,000+ patient records alongside clinical scheduling, lab integrations, and billing workflows. Approximately 800 employees including 200+ clinical staff who relied on the system daily.
Their 12-year-old patient management application — built on ASP.NET Web Forms with a monolithic SQL Server backend — had become a daily operational burden for clinical staff and a growing compliance liability for the network's leadership.
A 12-year-old system becoming a clinical liability
The patient management app had turned from a tool into a daily frustration — slowing clinical staff, blocking integrations, and failing HIPAA's evolving requirements.
Crippling Page Load Times
Screens averaged 6–8 seconds to load. Nurses spent approximately 45 minutes per shift just waiting — time that belonged at the bedside.
6–8 sec load timesDesktop-Only, Zero Mobility
No mobile or tablet access forced nurses to walk to fixed workstations to update patient records — a constant workflow interruption across every shift.
Workstation-only accessNo API Layer — All Manual Entry
Zero integration with lab systems, medical devices, and telehealth platforms meant every data point was entered by hand, introducing errors and delays.
0 system integrationsCompliance Report Took 2 Days
Monthly compliance reports required exporting to Excel and manually cross-referencing data — a full two days of work every single month.
2 days per reportHIPAA Compliance Gaps
The system lacked encryption at rest, comprehensive audit logging, and the security controls required by HIPAA's evolving requirements — a growing legal risk.
Encryption gapsNo Path to AI or ML
The hospital board wanted AI-powered patient risk scoring, but the legacy monolith had no data pipeline and no way to run ML inference — the architecture blocked every initiative.
No ML capabilityTwo phases, one cohesive transformation
We ran modernization and AI integration in overlapping phases — rebuilding the foundation while simultaneously building the intelligence layer on top of it.
- Re-architected monolith into API-first microservices using .NET 8, organized by clinical domain — patient records, scheduling, lab results, billing, reporting
- Responsive React front-end with mobile-first design — clinical staff now access the system from tablets at bedside
- Migrated to AWS RDS PostgreSQL — 200,000+ patient records moved with full audit trail and zero data loss
- HIPAA-compliant security — AES-256 encryption at rest, TLS 1.3 in transit, comprehensive audit logging, RBAC with MFA
- HL7 FHIR-compliant APIs enabling live integration with lab information system, pharmacy management, telehealth platform, and medical device gateway
- CI/CD pipeline with automated HIPAA compliance checks baked into every deployment workflow
- Patient data pipeline — aggregates vitals, lab results, medication history, visit frequency, and demographic factors into a normalized analytics store via Apache Airflow
- Gradient boosting risk scoring model (Python + scikit-learn) trained on 3 years of de-identified historical data — generates a 0–100 risk score per patient, updated daily
- Automatic high-risk flagging — patients scoring above 75 are surfaced with a visual indicator and physician alert, predicting 30-day readmission or adverse events
- Clinical insights panel explains each flag in plain language: missed medications, blood pressure trends, recent ER visits — not just a score
- SHAP-based model explainability — clinicians see exactly WHY a patient is flagged, enabling confident, informed action on every alert
Full-stack HIPAA-compliant architecture
Results that changed how they care
Measurable improvements across performance, compliance, clinical efficiency, and patient outcomes — all delivered within the 24-week engagement.
Page Load
Rate Reduction
Patients Detected
Report Generation
7 engineers, one clinical-grade delivery
A specialized team built for healthcare — combining HIPAA expertise, microservices architecture, and ML model training in a single cohesive engagement.
Solution Architect
System design, microservices decomposition, AWS architecture, HIPAA compliance strategy
Senior .NET Developer × 2
.NET 8 microservices, HL7 FHIR API layer, business logic migration, security implementation
Senior React Developer
Mobile-first clinical UI, bedside tablet interface, AI risk dashboard and insights panel
ML Engineer
Python, scikit-learn risk scoring model, SHAP explainability, Airflow data pipelines
DevOps Engineer
CI/CD, AWS ECS, automated HIPAA compliance checks, deployment automation
QA Engineer
Clinical workflow testing, data migration validation, security audit, zero-downtime rollout
24-Week Full Delivery
Zero downtime migration — 200,000+ records moved with full audit trail and no data incidents
What the client said
"The transformation i-verve delivered went far beyond a technology upgrade — it changed how our clinical teams operate every single day. Our nurses are no longer walking across the floor to update records, and our physicians are acting on real patient intelligence instead of gut instinct. The AI risk scoring has been a genuine clinical breakthrough: we've seen readmissions fall 18% in the first six months. This was healthcare IT done right."