2027 Summer Intern, MS/PhD, Software Engineer, Simulation Evaluation ML Model
About this role
The software engineering teams at Waymo build the intelligence behind fully autonomous driving technology, designing systems that allow vehicles to perceive surroundings, make split-second decisions, and navigate safely.
Incoming interns join the Simulation Evaluation ML Model team to design automated semantic extraction pipelines, develop temporal graph modeling techniques, and build multi-modal hybrid retrieval engines for failure analysis agents.
Graduate students with a strong foundation in Python or C++, deep learning frameworks, and a curiosity for autonomous systems will find an environment geared toward impactful research and mentorship.
What you will do
- Design and implement an automated semantic extraction pipeline that parses large-scale simulation logs, vehicle trajectory data, and multi-agent behavioral events into structured graph representations.
- Develop temporal graph modeling techniques to capture time-varying multi-agent interactions, road context, and safety-critical driving events.
- Build a multi-modal hybrid retrieval engine combining dense vector embeddings, keyword search, and graph traversal to enable fast scenario discovery.
- Create interactive data exploration tools and visualization dashboards to help autonomy engineers and researchers analyze complex scenario distributions.
- Benchmark retrieval precision, recall, and query latency against traditional tabular and relational search baselines.
- Collaborate cross-functionally with simulation researchers, machine learning engineers, and software infrastructure teams to document system architecture.
What you need
- Currently enrolled in a graduate program such as a PhD or Master's degree in Computer Science, Artificial Intelligence, Robotics, Electrical Engineering, or a related quantitative field.
- Must have at least one academic term remaining after the internship.
- Demonstrate strong software development experience using Python and/or C++ within a Linux development environment.
- Bring a solid foundation in core computer science concepts, data structures, algorithm complexity, and distributed data systems.
- Possess hands-on experience working with modern deep learning frameworks such as PyTorch, JAX, or TensorFlow.
Nice to have
- Demonstrated research background or practical experience in Knowledge Graphs, Graph Algorithms, Graph Neural Networks (GNNs), or Information Retrieval / Retrieval-Augmented Generation (RAG)
- Authorship of published papers in top-tier AI/ML, data mining, or computer vision conferences
- Experience with large-scale distributed data processing systems like Apache Spark, Apache Beam, distributed SQL query engines, or columnar data lakes
- Familiarity with temporal graphs, bi-temporal data modeling, or graph databases and tooling
- Demonstrated interest or domain experience in autonomous vehicle simulation, behavior prediction, motion planning, trajectory forecasting, or multi-agent interaction modeling
Benefits
- Competitive compensation packages with a housing/relocation bonus
- Medical, dental, and vision insurance
- Fun intern events and networking opportunities
- Free breakfast, lunch, dinner, and snacks
- Free access to Google shuttles and onsite gym
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 Mountain View, California.
What is the compensation for this role?
The expected hourly rate is $70 USD for Master's students and $85 USD for PhD candidates, with interns also eligible for a housing and relocation bonus.
What kind of technical background is required?
Candidates must be enrolled in a graduate program in a quantitative field, have strong development skills in Python or C++ on Linux, and possess hands-on experience with modern deep learning frameworks.
How does the hiring process work?
Resumes are accepted on a rolling basis until the role is filled, and applicants interested in multiple roles must apply to each one individually.