Interested in building a machine learning model to detect real faces from fakes? Curious about deploying ML models on the cloud to run at scale? Want to play with audio and video signals? If yes, we would love to learn more about you!
Who we are?
EYN is a London based startup that leverages cutting-edge computer vision, machine learning and cryptography to provide secure and fast online identity verification. EYN is supported by top VC investors like Entrepreneur First (backed by Reid Hoffman)
Our team has experience in developing world-leading biometric systems used in major airports and government agencies like JFK airport and FBI. EYN ’s team is made up of highly experienced individuals from some of the best academic institutions in the world like EPFL and Oxford.
Why are we building an identity security company?
We founded EYN with the mission to democratize secure identity verification to everyone, irrespective of any race, religion, colour or nationality. We believe in a future where identity verification is frictionless, effortless and without any human bias and judgment.
What does EYN do?
EYN provides an identity verification engine powered by border level technology that enables companies to verify their customer’s identity in seconds just by using a smartphone app. Currently working with the largest staffing companies in the UK, EYN’s software helps to onboard thousands of workers every month. Additionally. You can find out more about our products here: https://www.eyn.vision. You can also watch a product demo here: https://youtu.be/XiQ3bXlCQoo
The problem and EYN's approach
Registering users remotly requires checking both their identity documents and matching their facial photo to their selfie. Even though this approach offers big benefits to business to onboard users remotely, it can be subject to different attacks to spoof identities. These attacks consist of generating fake faces with images, videos or even masks. In order to detect fraudulent users, we generate ultrasound waves from the user's smartphone and based on the response of the audio waves we classify fakes faces from real ones- like a bat, you know!
- Training machine learning models usgin neural networks
- Testing machine learning models on debugging and production data
- Deploying ML models on the cloud (AWS)
- Write clean and well-documented code and APIs
- Provide training and support to internal teams
- Build reusable code and libraries for future use
- Proven work experience as a machine learning engineer (+ 3years)
- Experience or good understanding of neural networks
- Experience or good understanding of audio signal processing
- Experience in writing clean code, unit tests, and excellent debugging skills
- Good understanding of Python
Highly Regarded (but optional) :
- Experience with TensorFlow and Keras framework
What you will be working on?
- Review the current model for liveness detection
- Deploying model on the AWS cloud
- Train model using production data on the cloud
- MSc or PhD in Computer Science with machine learning and signal processing background.