Ready to Try Speechmatics?
Sign up for your free trial and we'll guide you through the implementation of our API. We pride ourselves on offering the best support for your business needs. If you have any questions, just ask.
Speechmatics hosted its very own Hackathon in a bid to get the team’s creative juices flowing before the holiday season. Hackamatics is a sprint-like event lasting three days whereby teams collaborate intensively over unique projects.
Traditionally, a Hackathon focuses on software projects, however, Hackamatics is open to all sorts of wonderful and bizarre endeavours both inside and outside of the software sphere and it doesn’t have to be speech related.
Hackamatics 2018 was a sequel to the very successful event last year whereby many projects were carried forward into 2018.
Some key projects from last year included Unomatics, our very own card game featuring the Speechmatics team, and a project that has become a proof of concept keyword extractor for media asset management. Hackamatics includes all the things you’d expect from an intense three-day Hackathon, including swag, burgers from Steak and Honour, burritos from Nanna Mexico, paninis, drinks, snacks, and cross-company fun and collaboration. Some projects coming from this years’ event included testing new tools and software to create greater process efficiencies, creating a company handbook for all things culture and processes and coming up with a Google Chrome extension providing real-time captions for all online video content rendered in the browser, amongst other cool projects.
We also had some fun projects this year such as a project to voice clone Stephen Fry to speak your own words using machine learning architectures in Pytorch. This project was particularly interesting with a graduate leading the way and attempting something completely new, working with the Applied Machine Learning team and learning whilst hacking this unique concept. One project used our speech technology on BBC News audio to play Pokemón through controller mapping. Finally, we made one of our Machine Learning Engineers a professional dancer/contortionist using machine learning and AI.
Sam Ringer, Machine Learning Engineer at Speechmatics said:
“Hackamatics was a three-day opportunity to do something really different with my work. I was able to risk using machine learning architectures that might not work and code using a variety of new tools. It was my first Hackamatics at Speechmatics and it was great fun working directly with other coders in the company outside my group. All the free food wasn’t bad either!”
I’d like to thank everyone for taking part and getting stuck into their projects. A lot of fun was had and it was great to see such collaborative teamwork to complete projects within the three-day limit. We’re looking forward to taking some of these creative projects forward into the new year.
Did you know that we’re hiring? Take a look at some of our roles here. Paul Aykroyd, Speechmatics