21Pictures represented an interesting challenge for the team at Shed Collective. Michael came to us with well documented research and rules and asked us to help build a proprietary matching algorithm that could essentially grade and match people based on numerous factors. We're well versed at building systems capable of profiling users for various applications such as Intern Avenue (matching interns and businesses) and Buckinghamshire Business First (matching content and services with interests). Drawing on this experience we delivered a proprietary algorithm and processing queue capable of matching across hundreds of thousands of users.
The biggest technical challenge we faced was to ensure the matching queue manager was capable of indexing and comparing hundreds of thousands of records without any significant delay or processing cost. This was achieved through careful planning, rapid prototyping and testing.