2027 Summer Intern, MS/PhD, Machine Learning Engineer
About this role
Waymo requires an advanced graduate student to train and fine-tune large multi-task transformers over driving-log sequences to improve evaluation quality for autonomous vehicle systems. Software engineers on this team build the brains of fully autonomous driving technology, enabling vehicles to perceive the environment and make safe decisions in complex scenarios.
During the internship, the selected candidate will design offline and end-to-end evaluations, build dataset pipelines, and write production-quality code in a shared codebase. Success in this role requires a strong background in deep learning fundamentals, hands-on experience with modern frameworks, and current enrollment in a graduate degree program focused on computer science or machine learning.
What you will do
- Train and fine-tune large multi-task transformers over driving-log sequences using JAX and Flax on TPUs.
- Iterate on fine-tuning strategies, training data mixtures, and loss functions to improve evaluation quality.
- Design and run rigorous offline and end-to-end evaluations including PR-AUC and calibration quality.
- Build the dataset and evaluation pipelines necessary for real hillclimbing A/B runs.
- Land production-quality code in a shared, high-traffic codebase.
- Communicate results through a design document, team deep dives, and a final intern presentation.
What you need
- Currently enrolled in a PhD or MS program in Computer Science, Machine Learning, or a related field, with plans to return to the program after the internship.
- Hands-on experience training and evaluating deep learning models in a modern framework such as JAX, PyTorch, or TensorFlow.
- Ability to build data pipelines, choose losses, and debug training runs effectively.
- Strong programming skills in C++ and Python.
- Solid grounding in machine learning fundamentals including precision and recall trade-offs, class imbalance, and experiment design.
Nice to have
- Authorship of published papers in top-tier AI/ML, data mining, or computer vision conferences like NeurIPS, ICML, ICLR, KDD, CVPR, CoRL, SIGMOD, VLDB, or ACL.
- Research or applied experience with transformer and sequence models, multi-task learning, transfer learning, and parameter-efficient fine-tuning.
- Experience with JAX/Flax, distributed training on TPUs or GPUs, and large-scale data processing pipelines.
- Familiarity with autonomous driving, robotics, or simulation.
- Familiarity with probability calibration, uncertainty quantification, importance sampling, active learning, or rare-event modeling.
Benefits
- Competitive compensation packages with a housing or relocation bonus if applicable
- Medical, dental, and vision insurance
- Fun intern events and networking opportunities
- Free breakfast, lunch, dinner, and snacks
- Free access to Google shuttles
- Onsite gym access
Skills and tech stack
Questions candidates ask
What is the remote policy for this internship?
This will be a hybrid onsite internship position located in San Francisco, CA.
What is the pay rate for the internship?
The expected hourly rate is $70 for Masters students and $85 for PhD students.
What tools and frameworks will I use?
You will work with JAX, Flax, TPUs, C++, and Python to train and evaluate large multi-task transformers.
How does the application process work?
Resumes are accepted on a rolling basis until the role is filled. Candidates interested in multiple roles must apply to each one individually, up to a top of three roles.