Jr. ML Engineer

BallerTV is a venture-backed startup building the world's largest sports network from the ground up. To date, our AI and computer vision video technology have enabled us to be the world's most prolific producer of live youth sports streams. We film games using a custom ultra-wide fisheye lens to capture an entire basketball court, soccer field, or volleyball court at once. Then we use machine learning to detect the playing area, the people, the balls, and anything else we need to completely understand the scene. We combine all of that information to create a trajectory for our virtual camera to follow the action similarly to how a videographer would film. Rather than relying on a physical actuator, we virtually move the camera by re-projecting the fisheye view. We then live stream the video anywhere in the world. And we do all of this in real time on an iPhone.

This is a junior-level role focusing on machine-learning projects related to sports and video processing. Strong skills in math, good spatio-temporal logic, experience with video preferred. Focus on optimizations is a plus. Strong coding and architecture skills desired.

Given the significant amount of creativity and brainstorming needed to build groundbreaking technology, this is an in-office role. Applicants must be available to work full-time on-site at our Pasadena headquarters.

Projects to include:

- Analyzing images to find playing surfaces and automatically bound and map them

- Analyzing video and images to detect players and the ball in a variety of sports using ML models developed in-house

- Analyzing video to detect events in a variety of sports

- Developing camera motion policies based on players, balls, and events

- Most models will run on iOS devices in real time, fully utilizing the CPU, GPU (Metal), and Apple Neural Engine (ANE), so optimizing for time is always important

- Working on CUDA GPUs

- Developing visualizations and introspection tools

- Working with an existing code base across Python, C++, and ML/DL libraries and also developing new code and architecture

- Working in a small team on substantial projects, with strict deadlines

- Contribute to design discussions

Skills to include:

- Strong Python knowledge

- PyTorch

- Git experience

- C++, basic knowledge necessary, but stronger skills highly desired

- Desire and willingness to learn and understand the projects

- Strong math and problem-solving skills

- Ability to reason about projects that depend on both space and time

Nice to haves:

- Experience with Metal, CUDA, Core ML 

- Knowledge of sports is welcome, but not a requirement; but must be willing to learn about sports to be able to develop projects related to them

Why BallerTV?

Competitive Salary

Equity Stock Options

Medical, Dental, and Vision Insurance

401(K) Contribution

Unlimited Vacation Policy

Immediate ownership and impact

We are an equal opportunity employer and all qualified applicants will receive consideration for employment without regard to race, color, religion, sex, national origin, disability status, protected veteran status, or any other characteristic protected by law.