MyEvaluations.com serves 10,000+ users across 900+ institutions: academic medical centers, community hospitals, military medical programs, and specialty boards. We are the engineering partner behind its AI features, modern clients, data-layer re-architecture, cloud infrastructure, and security posture.
Medical education software operates in one of the most demanding environments in enterprise SaaS. Every workflow must map to external standards: ACGME milestones, duty-hour limits, JCAHO procedure verification, and accreditation reporting. A single trainee accumulates years of evaluations, procedure logs, duty hours, and committee decisions, all feeding accreditation outcomes.
The overarching challenge: keep a mission-critical, deeply regulated, two-decade platform running flawlessly while simultaneously introducing defensible AI, delivering modern web and native mobile experiences, re-architecting the data layer, and holding an enterprise security posture that academic and military institutions can trust.
Traditional medical evaluations rely on numeric Likert scales where faculty rush through ratings and scores cluster, the opposite of what competency-based education needs. We built SAGE (Smart Adaptive Guided Evaluations), an AI evaluation engine that replaces static scale-based forms with a narrative-only, adaptive workflow.
On launch, SAGE reviews the trainee's prior 12 months of data in seconds, then generates tailored questions organized into competency-aligned sections. It asks adaptive follow-ups when responses are too thin to be useful, supports speech-to-text dictation on mobile, and auto-maps narratives to ACGME milestones. SAGE was presented at AAMC 2026 as an emerging teaching-and-learning technology.
Even with great narrative feedback, a Clinical Competency Committee still has to read hundreds of comments and turn them into defensible milestone decisions. We built two complementary AI layers:
We delivered a ground-up React rebuild that runs alongside the legacy application, migrating page by page via a strangler-fig approach with zero disruption to live programs. The modern stack includes Next.js 13.5, React 18, TypeScript 5, and Plasmic visual-builder integration. A design-system-first architecture with centralized tokens ensures consistency across the entire interface.
A production .NET 9 MAUI app (iOS and Android) with 600+ classes, MVVM architecture, and a shared UI component library. Clinical-grade features include geofencing and QR-code scanning for automatic duty-hour and conference attendance logging, and offline-first UX via SQLite and Realm for low-connectivity clinical environments.
The platform's SQL Server estate spans roughly 2,700 tables, 28,000 columns, 6,200 stored procedures, and 23 functional modules. We are executing a strangler-fig migration with a modern NestJS 10 + MikroORM 6 backend that coexists safely with the legacy .NET Framework application, plus a purpose-built optimization toolkit for ongoing stored-procedure rewrites, index optimization, and dead-code elimination.
AI-powered adaptive narrative evaluations with speech-to-text, milestone mapping, and competency-aligned questioning.
AI-driven narrative aggregation, competency summaries, predictive milestone scoring, and at-risk trainee detection.
ACGME duty-hour tracking with 80-hour and 1-in-7 rules, violation alerts, and mobile geofencing for automatic logging.
ACGME milestone mapping, semi-annual reporting, and Entrustable Professional Activity tracking across programs.
Procedure submission with supervisor verification, GAGES integration, and JCAHO procedure-verification compliance.
Attendance capture via QR-code scanning or attendance ID, with automated reporting and compliance tracking.
Specialty question banks for ITE prep, curriculum-linked assignments, and secure quiz delivery.
Compliance, accreditation, and affiliate-agreement document repository with centralized tracking.
Customizable surveys for ACGME compliance plus anonymous well-being screening tools meeting ACGME/LCME/CODA/CPME standards.
Standalone scheduler integrating up to 15 external scheduling sources including AMiON, QGenda, and Lightning Bolt.
White-labeled, secure onboarding for a specialized partner population with dedicated SSO and task management.
ERAS applicant import, SIMPL surgical assessment, multi-vendor scheduling, ACGME ADS export, and SAML SSO.
MyEvaluations is proof that a two-decade platform can simultaneously adopt defensible AI, gain modern web and native mobile clients, re-architect its data layer, run on modern cloud infrastructure, and hold an enterprise security posture, all without disrupting programs mid-academic-year.
Programs consolidate a dozen accreditation workflows onto one platform: one login, one source of truth, one audit trail. CCC preparation shifts from manual reading to review-and-refine with evidence pre-structured and milestone-mapped. And the platform sits at the center of each program's toolchain, eliminating duplicate data entry across applicant, scheduling, assessment, and accreditation systems.
Adaptive AI evaluations, summaries, and predictive milestone scoring
Modern web rebuild with Plasmic visual-builder integration
Native mobile with MVVM and offline-first UX
Modern backend coexisting with the legacy application
A 2,700-table estate under strangler-fig migration
Modern cloud infrastructure as code
Observability and error monitoring
Enterprise security with ACGME/JCAHO compliance
