Asset management

The Bank from Colart

A digital asset management solution which also provides optimised images and videos for the client’s standalone sites.

The client

A multinational manufacturer of fine-art products, including Winsor & Newton and Liquitex.

The brief

Organise the client’s assets into one place which can be accessed both internally and externally.

The outcome

Since this project began, the number of assets in the system has increased many times over and it provides assets to many external systems.

What we did

It was obvious early on with this project that the client’s assets needed a bit of organisation. As with lots of companies, there were various “silos” of assets which weren’t universally known about and lots and lots of duplication. So, the main priority was to gather all these and put them into logical groups based on file type and contents.

The client now has a system with nearly 100,000 assets covering images, videos, PDFs and other documents as well as ZIP files for collections of assets (such as DTP resources).

Storage solutions
We also decided very early on in the process that the asset files, along with the sized versions of images and videos would be stored in AWS’s S3 system. This allowed us to use a well-known API to get and put files into S3 as well as the ability to serve web-optimised images straight from there (via Cloudfront) to the client’s various web sites.

Opening the system to external users
The system became very successful as a hub for all the client’s assets. This led to a need to allow external agencies and photographers etc. to upload their work directly. Furthermore, uploading assets one by one was going to take too long, so we developed a bulk uploading system with a browser-based spreadsheet to ensure the assets’ metadata was correct. Also extra S3 buckets were created to allow certain external users to upload their files before adding the extra data.

Developing an API
Over time, a need to access the asset data and resources from different systems was identified. For this we developed a standalone API system which is heavily optimised to provide extremely quick response times. This is achieved through a mix of selective caching of data and requests and pre-processing the source data to make it as quick as possible to fulfill the request.