Senior / Staff Software Engineer, ML Datasets & Data Pipelines

Waabi Waabi · Robotics · Toronto, ON +2 · Remote · Autonomy & Algorithms

This role focuses on designing and implementing scalable ML data pipelines for training and evaluating deep learning models for an autonomous driving platform. It involves optimizing data formats, caching, dataloading, and improving data sampling for introspection and edge-case identification. The role requires strong software engineering fundamentals, Python proficiency, and experience with deep learning frameworks and distributed ETL pipelines.

What you'd actually do

  1. Design and implement data pipelines using real-world driving data and Waabi World (our high-fidelity simulator) to train and evaluate deep learning models.
  2. Optimize data formats, caching, and dataloading to drive highly efficient ML training and evaluation at scale.
  3. Improve data sampling and composition for deep data introspection to track model performance and uncover critical edge-case scenarios.
  4. Champion engineering excellence by writing high-quality, well-structured, and rigorously tested code.
  5. Help drive project roadmap planning, prioritization, and delivery.

Skills

Required

  • Python
  • deep learning frameworks (PyTorch, TensorFlow, JAX)
  • distributed ETL and data processing pipelines
  • ML pipelines (dataset management, dataloading, optimization)
  • cloud job orchestration, monitoring, and instrumentation

Nice to have

  • optimizing large scale distributed training pipelines
  • highly optimized ML inference pipelines
  • MapReduce (Apache Hadoop/Spark)
  • orchestration frameworks (Apache Airflow, Apache Beam, Google Cloud Dataflow, AWS Step Functions)
  • data challenges specific to autonomous driving
  • linear algebra (projections, transforms)
  • 3D geometry
  • multimodal sensor data (LiDAR, RADAR, camera)

What the JD emphasized

  • 4+ years of industry experience
  • deep learning frameworks
  • distributed ETL and data processing pipelines
  • dataset management, dataloading, and optimization
  • cloud job orchestration, monitoring, and instrumentation best practices

Other signals

  • design and implement data pipelines
  • train and evaluate deep learning models
  • optimize data formats, caching, and dataloading
  • improve data sampling and composition