This is an extremely rough and ready laravel project that has been created to provide a basic UI framework in support of a small coding challenge. The benefits of the simplicity of it is that it provides you with a lot of flexibility to focus on the problem, without being distracted by unfamiliar paradigms.
When recording diagnoses for patients, clinicians will typically pick from a short list, or will carry out a search to look for the diagnosis they want to record.
This project simulates this with a very basic Laravel application that has a Disorder model that simply contains the SNOMED code and the term for a couple of disorders. The index page provides an off the shelf search component that will perform an ajax search against this model through a simple route path.
Clinical terms are typically maintained in systems known as terminology services. A number of different implementations exist, all offering slightly different APIs. Currently OpenEyes maintains an internal list which users search against, but there is a desire to remove the need to maintain this internal list. Instead we want to be able to search against a 3rd party service through their API.
To achieve this, OpenEyes would need to provide a search endpoint that acts as an abstraction layer against whichever API endpoint has been made available for a given deployment.
A disorder that was found in a 3rd party system would still need to be stored in the application database. This supports referential integrity for any records that are linking to those disorders. Another challenge that therefore exists is for a disorder to be resolved in the backend if a disorder is selected that has not yet been stored in the application database.
The front end of this project is not relevant beyond simple demonstration of being able to perform a search. There's no value in making it pretty, and we are not concerned about configuration of the endpoint it is using.
The backend is what we particularly care about. There are a number of areas that are interesting and provide ample opportunity for you to demonstrate your skills.
- Testing
- Abstraction layers for the API
- Configuration for deployment
- Documentation
- Research regarding available terminology services
- Performance optimisation (through caching layers etc)
We don't expect you to spend days on this. We simply wanted to provide a defined problem space for you to work in, and provide a project around which we can have a follow up discussion. We'd recommend you thinking about 1 or 2 areas, and doing some work in that area.
We are interested in your ability to articulate and describe your approaches as much as the actual implementation. We do expect to see code in your submission, but descriptions of how you would take it forward are also of value.
We recognise that software development is undergoing some radical change at the moment. AI tooling is everywhere, and can make light work of a large amount of this. We are not averse to seeing what you might do using AI for this task, however it should be understood that this is first and foremost an opportunity for you to demonstrate your technical skill. If you use AI, we expect you to be clear what you have used it for, the approach you have taken with it, and a very clear understanding of the code that has been generated. We will also expect a much more complete solution, given the acceleration that the tools can provide.
This project was set up using Laravel Herd on Laravel 11. But you should be able to leverage the Laravel documentation to get started pretty readily.
Once you have cloned it, you can run
composer install
npm install
composer run dev
php artisan make:fresh --seed
npm run build
See here