Data Pipelines for Building Management

Client The Acadian Group (TAG)

Through our partnership with TAG we delivered the infrastructure for a cloud-based tool with a much larger potential—both in terms of scale and business impact—on its own capable of facilitating the construction and financial management of a multi-billion-dollar building project.

Mapping Software to Process

Massive-scale building projects today are often managed with little more than a large spreadsheet. Not only is this process prone to error, the tables it generates can get horridly complex, robbing users of the ability to view things at a higher level. And when your project budget is over a billion dollars, you've got a lot riding on your ability to avoid mistakes and see through mountains of data.

When TAG approached us, this was the problem they were attempting to solve: they were involved with a construction project that came with a whole lot of data, and had a whole lot at stake. 

Many data sources become one

TAG believed that, in order to better manage one's real estate assets, overseers needed to be able to view data from any given phase of a building’s life cycle: investment, construction, asset, or property management. There are a few legacy tools devoted to one phase or another, but none are universal. Any remaining gaps are most commonly filled by Excel, resulting in spreadsheets too large to be parsed by a human brain.

To answer the client’s immediate needs, we built a cloud-based ETL pipeline. It was designed to extract data asynchronously from the existing legacy sources, normalize it into a universal format, and then generate human-readable reports. This first phase was completed within five weeks, using a combination of Angular.js and .NET, and was immediately put to use by the team.

Making data useful

With the pipeline in place, we worked closely with TAG to extract more meaningful insights from the mountains of data coming in. We defined various classes of lifecycle- and role-specifc data views, and presented these visualizations using D3. These views were then tested with real users to make further refinements and improve our model. 

Something to build on

For the project’s remainder we continued to help our client sharpen its vision over workshops and conversations, engineered essential components, and shipped them to real people doing real work. And in the end, we delivered a core data pipeline and experience capable of managing massive-scale building projects, with an even greater future potential.