Welcome To GrayWorks Technologies

EverCare
Pediatrics

Category

Pediatric Home Health Care

Year

2026

Role

Design, Development, AI Integration & Hosting
EverCare Pediatrics

Project Description

EverCare Pediatrics LLC is a Georgia-based pediatric private home-care provider. Its Registered Nurses, Licensed Practical Nurses and caregivers deliver physician-directed skilled pediatric nursing, personal care, and companion & sitter services to children in their own homes — from tracheostomy, ventilator, G-tube and seizure care through bathing, feeding and mobility assistance to companionship and family respite — under the practice's own promise: Caring Today. Stronger Tomorrow.

GrayWorks designed and built the public website and the content system behind it. A warm brand palette and type scale drawn from the practice's own logo carry a home page, About, three service pages, Who We Serve, How Care Begins, Referrals, Careers, FAQ, Contact and the legal pages, with metadata, canonical URLs, a sitemap and structured data throughout. Every page, section, image, navigation item and setting is editable by staff from a built-in Next.js and SQLite admin with a rich-text editor and drag-and-drop section ordering, so the practice can change its own copy without touching code. All public copy came verbatim from the client's supplied material; anything unconfirmed shipped as a clearly marked placeholder rather than invented.

The site then grew into the practice's hiring system. A twelve-step public application wizard takes RN, LPN and CNA applicants through credentialing, electronic signature and document upload; an AI assistant reads the documents an applicant chooses to share and proposes form values for their review; and an AI phone agent calls applicants to agree an interview time, books it, and handles reschedule calls on the office line. Admins work the pipeline from an interview calendar, job projects with scored candidate matching, editable notification templates and a narrated in-app tutorial. The whole thing runs as a containerized deployment behind a TLS reverse proxy with nightly backups, covered by 27 Playwright end-to-end tests, runnable unit self-checks and a WCAG 2.2 AA contrast gate.

Key Deliverables

A staff-editable pediatric home-care website with an AI-assisted nurse recruiting pipeline built into it.

Client

EverCare Pediatrics LLC

Location

Decatur, Georgia — serving families statewide

Scope of Work

Everything GrayWorks delivered for EverCare Pediatrics:

  • Website Design & Brand System
  • Custom CMS & Admin Dashboard
  • Nurse Recruiting Pipeline
  • AI Document Assistant & Phone Agent
  • Interview Scheduling & Email Automation
  • Accessibility, Security & Hosting
Open Project

A warm, reassuring brand system built from the practice's own logo — a purple, orange and navy token palette with Nunito and Inter typography, defined once as design tokens and held in place by an automated contrast check. Public copy came verbatim from the client's supplied material across the home page, About, three service pages, Who We Serve, How Care Begins, Referrals, Careers, FAQ, Contact and the legal pages, with metadata, canonical URLs, sitemap, robots and structured data on every route. Staff edit all of it — pages, sections, media, navigation and settings — through a built-in admin with a Lexical rich-text editor and drag-and-drop section ordering.

A twelve-step public application wizard mirrors the practice's nurse credentialing packet for RN, LPN and CNA applicants: licence details with multi-state licensure rows, education, employment history, a pediatric and clinical skills grid, availability, references, credentialing documents, authorizations, and a drawn or typed electronic signature with a full audit trail. Each applicant receives a branded acknowledgement carrying their own Applicant ID, and the signed packet is rendered back as a faithful PDF. Admins get a searchable, filterable applications table with multi-select bundle PDFs, an internal review tab, one-time fourteen-day update-and-re-sign links with version history, and job projects that score approved applicants against each project's required skills.

Applicants may upload a resume, licence, BLS or PALS card and have an AI assistant propose values for the form, which they review field by field before anything reaches their draft. The model can only fill a fixed whitelist of fields, is instructed to refuse identifiers and health information, has its replies scrubbed server-side, and never touches the skills grid, signature or consent; documents read at that step are held in memory for the one request and never written to disk. A second AI agent works the phone: it calls applicants to agree an interview time and books it against the practice's own recurring availability windows, and it handles reschedule calls on the office line after verifying the caller by Applicant ID and surname. Both features fail closed — with no key configured, the site simply runs without them.

The site meets WCAG 2.2 AA, evidenced by an accessibility audit and held there by a token-contrast check that runs as a build gate. Contact and referral forms are schema-validated and guarded by a same-origin check, a honeypot and per-IP rate limits, and they explicitly ask visitors not to submit medical details — no clinical data is collected or stored, and no HIPAA compliance is claimed. Applicant documents are validated by magic bytes, written to a private volume that is never web-served, and released only through authenticated admin routes; admin logins use bcrypt hashing, login throttling and server-revocable sessions. It ships as a non-root container behind a shared TLS reverse proxy with security-header and rate-limit middlewares, a health check, a memory cap, nightly database backups, and a branded meeting subdomain that redirects to the practice's own video room.

Start a Similar Project

Your email address will not be published. Required fields are marked *