Healthcare

Patient Check-in and Cloud Engineering for a Telemedicine Platform

Online pre-check-in and self check-in apps for a Canadian imaging network, connected to the clinic RIS over HL7 FHIR R4, plus managed Azure, Terraform and CI/CD for its telemedicine platform.

10 weeks

From kickoff to the core pre-check-in and self check-in apps on Azure

FHIR R4

RIS integration, tested against the vendor’s production server

3 clinics

Visited in on-site discovery, 12 front-desk and imaging staff interviewed

Customer Story

A Canadian imaging network runs its own telemedicine platform: we built its patient check-in, connected it to the clinic RIS and moved the platform onto managed Azure.

A Canadian imaging network runs a patient–physician platform: after an exam, patients find their reports and images in their account, share them with their doctor and book video consultations. In 2024 we started work on that platform: first a clinic workflow study and the check-in apps, then infrastructure, CI/CD and QA inside the client’s product team.

Healthcare

Industry

Custom Software Development, Healthcare Integrations, Cloud & DevOps, Team Extension

Services

Canada

Location

Custom software · Cloud & DevOps · Team extension

Engagement

Challenge

Overloaded front desks and a platform that had outgrown its first architecture

Front desks were overloaded. In our on-site study, walk-in patients shared one queue with booked patients and took several times longer to register, and technologists filled in intake questionnaires by hand before first exams. Patients repeated the same details and paperwork at every visit. The platform had outgrown its first architecture: front end and back end were tightly coupled, the database ran on a single virtual machine and video calls ran on fixed servers. Every change had to meet Canadian health-privacy rules and work with the clinics’ existing RIS through a vendor API that was still being built.

Solution

Pre-check-in, self check-in and RIS integration on Azure

We built pre-check-in so patients can check in and complete intake questionnaires on their phone before the visit, added a self check-in app for the clinic, and connected both to the clinic RIS over HL7 FHIR R4. In parallel we moved the platform’s database and deployments to managed Azure with Terraform and GitLab CI/CD, and later joined the client’s product team.

Illustration of three phone screens: signing in with an MRN or health card number, checking personal details with an intake questionnaire progress bar, and an exam ticket with a QR code (sample data)
Pre-check-in on the patient's phone: sign in, check details, get a QR ticket
Illustration of a clinic check-in tablet showing the consent step for a mammography exam, a short educational video and a 'You're checked in' confirmation (sample data)
Self check-in at the clinic: QR or MRN sign-in, consent and a short exam video
Illustrated architecture diagram: patient phone and clinic kiosk through Azure Front Door to the check-in apps in Azure Canada Central, linked to the clinic RIS over HL7 FHIR R4; below, managed Azure database, Jitsi on Azure Kubernetes Service, GitLab CI/CD and Terraform
Architecture: check-in apps on Azure, HL7 FHIR R4 to the RIS, platform DevOps

What we delivered

Discovery in the clinics

We visited three clinics and interviewed receptionists and mammography and ultrasound technologists, mapping each step from arrival to report. The study produced a prioritised backlog: move check-in and intake questionnaires out of the waiting room and onto the patient’s phone.

Pre-check-in

Before the visit, patients get an email link, sign in with their MRN or health card number, check and correct their details, confirm or cancel, and complete the intake questionnaire for mammography or bone density — resuming from the same link if they stop. They receive a boarding-pass style ticket with a QR code, a calendar file and directions, in English or French.

Self check-in at the clinic

A self check-in app signs patients in by QR code or MRN, handles walk-ins, records procedure-specific consent, shows short educational cartoons about the exam and confirms they are checked in.

RIS integration over HL7 FHIR R4

We integrated both apps with the clinics’ RIS through its HL7 FHIR R4 API — appointment lookup, MRN search, patient updates, confirmation and cancellation — with an hourly cache for when the RIS is slow. The integration passed testing against the vendor’s production FHIR server in January 2025.

Cloud and delivery pipeline

We instrumented and load-tested the platform API, moved its database from a virtual machine to a managed Azure database with integrity checks, cleaned up the API’s HTTP methods, wrote the production environment in Terraform and set up GitLab CI/CD with linting, SonarCloud checks, automatic deploys to dev and one-click releases to QA and production.

Video calls on Kubernetes

We moved the Jitsi video stack into containers on Azure Kubernetes Service with scaling rules, so call capacity is no longer fixed by one server.

Security for the check-in apps

Hosted in Azure’s Canada Central region behind Azure Front Door, with secure login configuration, encryption in transit and at rest, intrusion detection, and patient data cleared from the apps when a session ends.

Inside the product team

From September 2024 our tech lead, QA engineer and DevOps engineer worked inside the client’s product team, under its management, on QA, CI/CD and Azure infrastructure. Our DevOps engineer still supports the platform in 2026. We also built the platform’s AI symptom triage assistant.

Tech Stack

The stack we worked in: Azure, Terraform and Kubernetes

Azure Kubernetes Service, Front Door, Container Registry, Communication Services and a managed database on Microsoft Azure, Terraform for the production environment, GitLab CI/CD with SonarCloud checks, Docker, Jitsi for video, and HL7 FHIR R4 for the RIS. The client’s platform itself runs on Laravel, React and an OHIF-based DICOM viewer.

Microsoft Azure
Kubernetes
Terraform
GitLab CI/CD
SonarCloud
Docker
Jitsi
Roadmap

Project duration

From clinic discovery in February 2024 to DevOps support in 2026.

Feb 2024

Clinic Discovery

On-site discovery in three clinics; workflow and inefficiency report.

Mar–May 2024

Check-in Apps

Pre-check-in and self check-in apps on Azure, with QR ticket, calendar file and directions.

Jun–Aug 2024

French Version & Cloud

French version, bone-density questionnaire, questionnaire resume; load testing, database to managed Azure, Terraform and GitLab CI/CD; Jitsi to Azure Kubernetes Service.

Sep 2024 →

Team Extension

Tech lead, QA and DevOps engineers join the client’s product team.

Nov–Jan 2024–2025

RIS Integration

HL7 FHIR R4 integration with the RIS, tested against the vendor’s production server.

2025 → 2026

AI Assistant & Support

AI symptom triage assistant; QA engineer (to April 2026) and DevOps engineer (ongoing) inside the client’s team.

Team Size

1 × Project Manager
1 × Tech Lead
1 × PhD Solution Architect
1 × Front-end Developer
1 × Senior Back-end Developer
1 × DevOps Engineer
1 × QA Engineer

What the Engagement Delivered

Delivery milestones from the project records. Our engineers have worked inside the client’s product team since 2024, and DevOps support continues in 2026.

10 wks

Core check-in apps

Core pre-check-in and self check-in apps delivered about ten weeks after kickoff (May 2024), with the RIS mocked until the vendor’s API was ready; French, the bone-density questionnaire and questionnaire resume added by August 2024.

FHIR R4

RIS integration

RIS integration over HL7 FHIR R4 completed and tested against the vendor’s production server in January 2025.

Azure

Managed cloud and CI/CD

Platform database moved from a VM to a managed Azure service, production described in Terraform with automated GitLab deployments (July 2024); Jitsi video moved to Azure Kubernetes Service (August 2024).

FAQ

Patient check-in and platform engineering: common questions

Our core pre-check-in and self check-in apps took about ten weeks, with the clinic system mocked. Connecting to the RIS took longer because it depended on the vendor’s API, so plan the timeline around the integration, not the screens.

Yes, if the RIS exposes an API. We used the RIS’s HL7 FHIR R4 API to look up appointments, search by MRN, update patient details and confirm or cancel visits, with an hourly cache for speed.

We moved the Jitsi video stack into containers on Azure Kubernetes Service with scaling rules, so call capacity grows with the number of calls instead of being fixed by one server. Recording needs its own scaling plan.

The check-in apps were built for PIPEDA requirements: hosted in Azure’s Canada Central region behind Azure Front Door, encryption in transit and at rest, procedure-specific consent, and patient data cleared when a session ends. Formal compliance sign-off stays with the clinic.

Yes. From late 2024 our tech lead, QA and DevOps engineers worked inside the client’s own team and backlog, under the client’s management; our DevOps engineer still supports the platform today.

Discover More Related Customer Stories