Curated technical proof artifact for MealMind, a Wilds, Inc. patient meal ordering application.
This repository is intentionally not the production source tree. It exists to show the thinking, architecture, product judgment, and representative code behind Justin Wilds' first serious application build without exposing the turnkey product, private integration details, deployment configuration, or patient-facing implementation surface.
Start here if you are evaluating Justin Wilds or MealMind technically:
- Read WHAT_THIS_PROVES.md for the short evaluator view.
- Read ARCHITECTURE.md for the system shape and integration boundary.
- Read ADR_DIGEST.md for decision discipline.
- Read SECURITY_AND_PHI.md for the PHI boundary and safety posture.
- Read VOICE_AND_AI.md for AI and voice integration posture.
- Read OPERATIONS.md for the patient and operator workflows.
- Read CODE_EXCERPTS.md and browse
code-samples/for implementation style. - Read PUBLIC_BOUNDARY.md before asking why production internals are not here.
MealMind is a patient-facing meal ordering layer designed for hospital environments:
- browse-and-tap meal ordering
- patient-appropriate authentication
- multilingual menu presentation
- voice ordering as an optional enhancement
- bedside TV / companion display concepts
- host dietary-system integration boundary
- PHI-minimizing session design
- accessible, low-friction patient UX
The product question was: how can a hospital patient order food without fighting a legacy text-heavy interface, while still respecting diet, allergy, security, and operational constraints?
The production system is a full-stack TypeScript application:
React + TypeScript client
Express + Node backend
SQL Server persistence
host dietary-system adapter boundary
voice and AI services behind optional feature surfaces
Azure deployment target
The proof repo does not include the full source tree. It includes curated documentation and sanitized excerpts.
| File | Purpose |
|---|---|
WHAT_THIS_PROVES.md |
Short evaluator guide: what the repo is evidence for |
ARCHITECTURE.md |
System shape, data flow, integration boundary |
ADR_DIGEST.md |
Selected decisions and why they matter |
VOICE_AND_AI.md |
Voice ordering and AI posture without vendor lock-in |
SECURITY_AND_PHI.md |
PHI boundary, session, audit, and safety model |
OPERATIONS.md |
Patient, guest, display, and staff workflows |
CODE_EXCERPTS.md |
Guide to the included representative samples |
code-samples/ |
Sanitized implementation excerpts |
PUBLIC_BOUNDARY.md |
What this proof repo intentionally excludes |
PROOF_REPO_STANDARD.md |
Reusable standard for Wilds public proof repos |
This proof repo does not include:
- full application source
- production Git history
- deployment workflows
- environment files or real resource names
- private host-system contracts
- real patient, facility, or menu data
- enough schema or code to recreate the product
That boundary is intentional. The repo proves the work without giving away the asset.
- Patient-safe UX design under healthcare constraints.
- PHI-minimizing architecture.
- Adapter-first integration thinking.
- Voice and AI used as optional access layers, not brittle dependencies.
- Security decisions captured as architecture, not bolted on later.
- Strong documentation and ADR practice from the first app onward.
Curated proof artifact. Production MealMind remains private.