About MAGIC AI
MAGIC AI is an elegant in-home health coach that utilises computer vision, connected weights, and cameras to provide personalised training sessions led by world-renowned athletes. We have gained recognition and exposure, being stocked in Selfridges, featured on Good Morning Britain, and listed as one of Fast Company's World's Most Innovative Companies of 2024. With just a small team, our company achieved a substantial revenue rate during its first financial year.
MAGIC AI is the pioneer in offering users the opportunity to perform exercises in front of an intelligent hologram mirror. Our cutting-edge proprietary computer vision software enables real-time form correction, rep counting, and live feedback, resulting in a fully customised workout experience. We have also collaborated with top athletes to create exclusive workout content, revolutionising the way users can be trained in sports, strength training, yoga, dance, and more. With us, individuals can benefit from personalised tracking and feedback, as if being privately coached and corrected by the world's best, without the hefty fees typically associated with personal training.
We have secured venture capital funding from prominent multi-billion-dollar VC funds, who have also invested in well-known companies such as Grover, Scalapay, Onfido, as well as notable founders and angels from Spotify, Stripe, Facebook, Hopin, and Tough Mudder.
About the Role
You will be joining our Data Science team working on our powerful ReflectAI® tracking technology, which enables real-time form correction, rep counting, and live feedback across modalities such as strength, cardio, yoga and pilates. As also member of our IT team you will be reporting the our CTO. Our small team consists of highly accomplished individuals, each having achieved prior exits. You will find this role to be highly fulfilling.
What is Involved
- Collaborate closely with other Data scientists and our Engineering Team to advance our groundbreaking ReflectAI® technology.
- Work extensively with MediaPipe and MoveNet to leverage pose estimation coordinates provided by the libraries.
- Develop rules/algorithms to accurately calculate metrics such as total reps, rep progress, rep speed and form quality by analysing the outputted coordinations.
- Build and train new AI models
- Optimise AI models to run on specific hardware accelerators
- Utilise statistical analysis techniques to minimise inaccuracies when dealing with complex occlusion scenarios.
- Collaborate with the Product team to ensure technical and product requirements are understood and executed on.
- Work closely with QA to resolve bugs and minimise any bugs that appear in staging/production releases.
- Ensure smooth and regular release of new features including coordination and communication with Customer Support and Marketing.
Requirements
- Building and training new AI models
- Experience with feature extraction from time-series or sequential data.
- Ability to develop and implement algorithms for real-time data processing and analysis.
- Strong foundation in linear algebra, geometry, and basic calculus.
- Good understanding of probability and statistical analysis.
- Good communication abilities, particularly in effectively and clearly explaining technical concepts to non-technical individuals and in writing detailed project documentation.
- Meticulous and thorough in paying attention to details.
- Self-motivated and capable of working independently.
- Able to work effectively and cooperatively in a team setting.
- You don't have to be a regular gym-goer, but you should be passionate about developing technology that will revolutionise how people exercise!
Nice to Haves
- Experience using MediaPipe, MoveNet or other machine learning models, ideally pose estimation.
- Experience coding in Kotlin, Java or C++.
- Proficiency in using JIRA or other project management tools.
- Experience collaborating in Agile Scrum teams.
- Experience as a data scientist within a startup / high growth environment.
- Knowledge of correct technique across strength exercises (e.g. squat, deadlift) and other fitness related modalities.
- Product minded with a good understanding of how to build products.
- Passion for fitness, ideally with experience building fitness products and working out.
Benefits
- Competitive salary.
- Share Options in the company.
- An impact from day one. Our business is scaling by the day. You'll work on ambitious projects, and your contribution will significantly impact the success of MAGIC AI now and in the future.
- Unlimited Holiday (self-directed time off)
- Flexible Home/Hybrid Working from our London HQ (At least 2 days WFH per week)
- Mental Health Wellbeing support
- Hardware budget for brand new Macbook or other.
- Professional learning & development budget.
- All. The. Fun. Regular awesome socials.
Top Skills
What We Do
MAGIC uses computer vision, connected weights and 3D cameras to deliver personal training by the world's best athletes at home. MAGIC is the first in the world to allow users to perform exercises in front of an intelligent hologram mirror which uses our own custom computer vision software to correct real-time form, rep count, and give live feedback to give a completely hyper-personalised workout. We have combined this with our patent-pending dumbbells and co-created workout content with the world's leading athletes. Similar to how Masterclass.com has done for the higher arts, we enable anyone to be coached by the world’s best in sport, strength, yoga, boxing, dance etc. Users get 1:1 hyper-personalised tracking and feedback as if being personally trained + corrected by them -- but without having to pay £100s a week in typical personal training fees. We are stocked in Selfridges, were named Techround’s #1 IoT startup 2021 and were recently backed by Virgin Startup. We are venture capital backed by 2 leading multi $billion VC funds who also invested in unicorns: Grover, Scalapay and Onfido as well as the founders/angels from Spotify, Stripe, Facebook, Hopin and Tough Mudder.
