CrowdScores

Software Development at CrowdScores

We strive to use the best practice tools, processes and architecture to produce unique software that is relied on by hundreds of thousands of users daily. Our developers have the freedom to choose and update the libraries and tools used to suit them and the problems we are trying to solve. Through code reviews and retrospectives we aim to continually inspect and improve our methods and to learn from each other.

We make our development process as rapid as possible, using microservices on the backend and Clean architecture in the mobile clients. Scrum, Continuous Integration and extensive unit tests allow multiple small releases dailly (pushed to non-beta users weekly on mobile).

We build for scalability, allowing us to cope rapidly with ten fold increases in demand at peak times.

Our backend services process millions of Tweets using Natural Language Processing with the potential to use machine learning in a coming overhaul.

What We Do

CrowdScores provides live football data via a unique method: crowdsourcing.

Our app empowers loyal fans, no matter which team they support, to share the latest events live from the match. Every report goes into our processors which employ complex algorithms to determine inaccurate or malicious reports. Within milliseconds our systems can determine the veracity of a report and update the score for everyone using the CrowdScores app, performing at scale even when tens of thousands of users are following games.

Our backend software uses a microservice architecture, allowing us to deploy and scale rapidly to cover the biggest events in football. We are now expanding our service to cover amateur football, with hundreds of 5- and 7-a-side leagues in the UK using the app to keep score, letting players see their data alongside the pros.

The CrowdScores Story

CrowdScores is a social scores network which allows fans to enter an immersive football experience where they can share both in-match events and opinion.

Born out of a frustration at a data monopoly which ignores numerous levels of the footballing pyramid and offers a slow service by modern technological standards, CrowdScores was formed in January 2012 with a vision to redress the balance of football data coverage.

Initially setting out to revolutionise how fans share football events, we soon realised that by crowdsourcing live football data it increased the speed and quality of the information. This led to the development of the CrowdScores app, a dedicated platform for fans to share in-match events so that we are faster, more comprehensive and more accurate than any established organisation.

CrowdScores’ first steps were made with a team of just two before expanding rapidly to reflect an improving system and ever growing user numbers. Now a team of more than 20 and with a user base of hundreds of thousands, CrowdScores continues to break ground in the world of sports data.

Where We Are

CrowdScores is based in central London and we comprise of a team from around the world.