LIV Residential

Helping people find their new home reliably, transparently, and securely, all over The Netherlands.

Purpose

LIV Residential needed a custom-built CMS to manage their property portfolio, and required a tailormade frontend to display the data on for their customers. It required location-based filtration tools and comprehensive visual and legible information relevant to prospective renters or buyers.

Responsibilities

For this project, I acted as a floater between backend, frontend and management. I led the project internally, being part of communications with the client, guided and helped develop frontend features, and assisted on expanding the content management needs.

From a project perspective, we used Redmine, a lightweight waterfall-style project management tool. We had initial designs approved before my start on the project, and from that tickets were created and added to the fold, from project initializations to features.

Due to the nature of our project, we avoided applying INVEST principles when writing tickets. We believed that a concise and clear list of to-do's and use cases would make it so that any young employee could jump on the project. Scrum was not a methodology active at this stage in the company yet, outside of daily standups.

From a frontend perspective, we primarily used PhalconPHP's View system, alongside SCSS style transpiling and Docker for hosting the images to our local environments, so that we don't run into any device-related issues. This was a specialty in the company, and there was no need to apply a dynamic framework like React or Vue to a frontend like this. Speed was of the essence.

We created loose components so that we could apply these to our fetch outputs from our controllers. These components on the client-side would allow for each page to look and feel uniquely like LIV, with the exception of the homepage. We also used CSS variables for themeing.

My responsibilities were tied towards creating the filtration mechanisms, alongside some minor component work here and there. The search page had a mobile-friendly system, with a collapsable filtration menu, and a results page. Eventually an emerging feature became to add a Map View, to model other sites like Pararius and Funda, the two biggest real estate platforms in The Netherlands. We used Google Maps' package and plugged it in similar to what we did for the King Penguin AR game, so a lot of code was reused and styling was adjusted. We did create a click component to view a smaller box version of the listing before users navigate towards it.

Overall the project took around 7 months to deliver the first Alpha for review. The CMS was already largely generated through the upload of an SQL database architecture diagram, so most of the backend was already ready to go. The construction of the components didn't take long and the filtration functionality went smoothly.

What I would do differently is adopt a different project methodology. The client was left in the dark throughout the process, and as such when feedback did eventually come, the project scope ballooned past my tenure. But in the end the project did go live, though it did undergo visual renewal since then.