Project Study Case
The Problem
Integra is an hourly rental clinic for healthcare professionals, helping them save time and money managing rooms. Their scheduling was mostly manual in spreadsheets, which worked for a single space but broke down across multiple locations and booking states (paid, executed, cancelled, etc.). They needed to avoid human mistakes, double-bookings, and time wasted in back-and-forth communication with professionals and backoffice.
The Proposal
An integrated web app to manage scheduling, payments, professional registers and space availability through two main views:
- Backoffice (Recepcionists) vision → shows all spaces and their schedules, letting staff create, edit, cancel, and mark executions and payments, as well as manage professionals in the database.
- Healthcare Professional Vision → shows available time slots across spaces without exposing other professionals’ details, allowing users to book, cancel, reschedule, and review their payments and past appointments.
The Architecture
this project needs to be fast and with good UX because the professionals need to focus in their patients and not in managing a scheduling software. Based on that we decide to opt for the following technologies:
MongoDB was the choice for databases for fast-schema flexiblity and faster queries for time based data, with aggregation pipelines to solve join situations.
Golang was the choice for generating backend for being a solid, time-proof language that has good performance and updates that are retrocompatible, unlike Node.js distributions.
Not only REST but Websockets was needed so the users can schedule and see data in realtime, this prevents race-conditions and gives a better experience to the users, as they can see available slots in realtime.
Svelte was the choice for frontend for being a reactive framework that is very fast, good DX, performant and very good to integrate with realtime application, mostly chosen for fast-delivery (less boilerplates) frontends
Solution Highlights
One view for everything.
Users can manage appointments, payments, and schedules in a single view, avoiding page switching and cognitive overhead, preventing lines and communication with professionals without loading breaks.
The technique for making this is:
- Drag and Drops to prevent cognitive overhead
- All the interactions except payment can be done by mouse
- Right-click menu to solve edge cases easily
- Three-clicks to anything.
Performance: 400ms to populate the whole system.
The system is made to prevent as much as possible all the loading screens. I did that by making requests in parallel on the client, with Goroutines and fasthttp in Go. Insert and Update operations can take 200-300ms. Everything can be processed in less than 1s. thanks to Golang performance and MongoDB fast list, insert & update operations that performs in nanosseconds to 1ms, all the time took is just download, unzip (from gzip) and read the data.
Maintenance Heaven
Everything is documented. Storybook for components and OAS (OpenAPI Specification) alongside with Stoplight Elements for managing and retrieving all the API Documentation, with edge-cases to provide top-notch documentation for an amazing and future-proof DX.
One of the most important things for this project is to be maintainable because as an ERP, companies changes all the time, from simple features from strategic decisions, every decision can also be changed in the software.
Results
No lines, no chaos.
- Eliminated waiting lines caused by scheduling errors and delays.
- Removed front-desk bottlenecks by enabling instant scheduling and updates.
- Cut unnecessary back-and-forth communication to near zero.
Bulletproof Scheduling
- Zero double-bookings, even during peak usage.
- Real-time availability guarantees every visible slot is valid and bookable.
- Clear and enforced booking states eliminated operational and financial inconsistencies.
From Minutes to Seconds
- Zero double-bookings, even during peak usage.
- Real-time availability guarantees every visible slot is valid and bookable.
- Clear and enforced booking states eliminated operational and financial inconsistencies.
Real Impact on the Ground
- Professionals no longer depend on reception, they control their own schedules.
- Patients move through the clinic faster, with fewer delays and friction.
- The clinic operates with a predictable, continuous flow instead of interruptions and queues.
Future Improvements
The next step is to evolve the platform from a scheduling tool into a daily operating system for professionals - and eventually, an intelligent system that reduces manual work to near zero.
Professional Experience Upgrade (Improves UX)
Transform the product into a tool professionals rely on daily, not just occasionally.
- Personal analytics dashboard with revenue (weekly/monthly) and occupancy rate
- One-click rebooking of past schedules
- Calendar sync (Google Calendar, iCal)
- Mobile-first experience / PWA
The goal is to increase engagement and make the platform part of the professional’s daily workflow.
Intelligent Scheduling Layer (AI Integration)
Move beyond traditional interfaces by introducing intent-based interactions and automation.
- AI-Assistant for Receptionists: A real-time assistant to executing actions and answer questions
- Categorization & Auto-tagging of patterns in appointments.
- Automate Dashboard creation charts with AI.
- AI/Statistic chance of no-show or cancelling.
The goal is to reduce manual work to use AI to automate tasks and provide insights.