Senior Software Engineer – ML & ADAS Perception

Job Summary

Join our dynamic team as a Senior Software Engineer specializing in ML and ADAS Perception, where you will develop cutting-edge machine learning solutions that enhance autonomous driving and advanced driver-assistance systems. This role offers an exciting opportunity to work with large-scale sensor and video data, build robust perception algorithms, and contribute to state-of-the-art simulation frameworks. You will collaborate with cross-functional teams to drive innovation in autonomous vehicle perception, applying your expertise in deep learning, computer vision, and software engineering in a fast-paced, on-site environment in Bangalore.

Key Responsibilities

• Develop and maintain machine learning-based solutions for ADAS/Autonomous Vehicle (AV) perception and data-driven simulation pipelines.
• Train, fine-tune, and optimize deep learning models using extensive ADAS/AV datasets to improve perception accuracy.
• Design and implement algorithms for object detection, multi-object tracking, segmentation, and depth estimation to enable reliable vehicle perception.
• Apply advanced state estimation techniques, including ADAS tracking and filtering algorithms such as Kalman filters, to enhance model robustness.
• Build and automate ML pipelines for simulation artefact generation and data processing, ensuring reproducibility and efficiency.
• Work extensively with camera geometry, including intrinsic/extrinsic parameters, lens distortion correction, and camera projection modeling.
• Develop validation frameworks, testing suites, and metrics to rigorously evaluate ML pipeline outputs and perception algorithm performance.
• Analyze model outputs and pipeline results rigorously, providing feedback loops for continuous improvement and retraining of models.
• Handle, preprocess, and analyze large-scale video and multi-sensor datasets using Python-based tools to support perception workflows.
• Debug and troubleshoot complex issues in ML pipelines, perception algorithms, and related software components.
• Collaborate effectively with cross-disciplinary teams to deliver reliable and scalable ADAS/AV perception solutions.
• Adhere to software engineering best practices including version control with Git, Linux-based development, test automation, and continuous integration.

Required Skills and Qualifications

• Bachelor’s degree in B.Tech, BE, or equivalent technical discipline.
• Minimum of 2+ years of professional experience in Machine Learning, Computer Vision, ADAS, or Artificial Intelligence domains.
• Strong programming proficiency in Python with hands-on experience in PyTorch or equivalent deep learning frameworks.
• Solid understanding of Computer Vision concepts and Image Processing techniques.
• Proven experience with object detection, multi-object tracking, segmentation, and depth estimation model development and deployment.
• In-depth knowledge of camera geometry including intrinsics, extrinsics, lens distortion, and projection models.
• Experience in model training workflows, fine-tuning, dataset management, data labeling, and model evaluation methodologies.
• Foundational knowledge of linear algebra, probability, and state estimation techniques relevant to tracking and filtering algorithms.
• Practical understanding and application of Kalman filters and related tracking algorithms.
• Proficient working knowledge of Linux OS and common development tools.
• Experience with Git and version control systems.
• Excellent analytical, problem-solving, debugging, and communication skills.

Preferred Qualifications

• Experience with C++ programming in addition to Python.
• Hands-on experience with NumPy, pandas, and data visualization libraries for exploratory data analysis and reporting.
• Familiarity with automated testing frameworks such as pytest and experience implementing regression testing.
• Understanding of perception evaluation metrics including Precision, Recall, Intersection over Union (IoU), Multiple Object Tracking Accuracy (MOTA), and Higher Order Tracking Accuracy (HOTA).
• Exposure to ADAS/AV simulation environments and standards such as ASAM OpenDRIVE and OpenSCENARIO.
• Experience with simulators like CARLA or esmini for data-driven simulation and scenario generation.
• Prior work with large-scale video, sensor, or vehicle datasets in Python-based processing pipelines.