Netflix is looking to improve the performance, reliability, and efficiency of its infrastructure systems by forecasting traffic and demand, and building models to optimize capacity and traffic steering decisions.
Requirements
- Experienced in developing and implementing machine learning models with a successful track record of driving business impact.
- Deeply familiar with the ML lifecycle and strong technical judgment when assessing different solutions for deploying models in production.
- Experienced in and motivated by the infrastructure domain, having worked on large distributed infra systems on topics such as demand forecasting, capacity planning, traffic steering, load balancing, etc.
- A strong coder with experience in Python and standard ML frameworks like PyTorch and TensorFlow.
- Experience with languages like Java or C++ is a plus.
- Familiarity with optimization models with standard frameworks/solvers (e.g., XPress, cvxpy, Gurobi) is a plus.
- Familiarity with operational tooling for ML services (monitoring, alerting, etc.), and services for model hosting/serving.
Responsibilities
- Forecast and predict key business and technical inputs such as traffic volume and resource demand across our fleet of cloud services and systems
- Build, update and maintain machine learning models to optimize our infrastructure footprint on topics such as capacity planning, autoscaling, loadshedding, and traffic steering
- Own the end-to-end model lifecycle including ideation, feature building, training, evaluation, monitoring and continuous improvement
- Partner with software engineers to identify high value opportunities to apply modeling techniques to improve the performance of infrastructure management systems
- Partner with ML engineers and data scientists on model observability initiatives and experimentation to improve model design
Other
- An exceptional thought partner with strong communication skills, able to explain complex technical concepts clearly to cross-functional partners and business leaders
- Comfortable with ambiguity, with a strong ownership mindset, and thrive with minimal oversight and process
- Live Netflix values while bringing a new perspective to continue improving our culture
