Lead Perception Engineer

We are an early-stage, maritime defense technology startup backed by top tier investors. Our team is designing and building autonomous and highly producible ships that can serve multiple missions on the open ocean.

We are seeking a Lead Perception Engineer based out of our Lexington, MA headquarters with hybrid/field-based options to architect, implement, and operationalize the end-to-end perception, sensor ingestion, and sensor fusion pipeline for autonomous maritime navigation. This is a foundational, hands-on technical leadership role
. You will personally build the initial perception and data ingestion stack while establishing the technical direction and standards for a future perception team. The ideal candidate is equally comfortable designing system architecture, writing performance-critical code, and working in the field to ensure sensors and data pipelines operate reliably in real conditions. You will work closely with hardware, autonomy, and systems engineering teams to ensure sensor choices, data ingestion, and perception outputs scale with the company’s roadmap.

Autonomy at sea introduces unique perception challenges—long-range sensing, dense and ambiguous scenes, dynamic backgrounds, adverse weather, and limited infrastructure. Solving these problems requires engineers who can build robust, production-ready perception systems
, not just prototypes.

U.S. Person status is required as this position needs to access export controlled data.

• Design and implement the full perception stack for autonomous maritime navigation, including:
• Sensor fusion across vision, navigation sensors, and future modalities
• Interfaces to planning, control, and autonomy systems
• Reliable, time-synchronized ingestion of high-bandwidth sensor streams
• Scalable abstractions that support evolving sensor configurations
• Data formats and APIs that enable downstream perception, logging, and replay
• Define system-level requirements, performance metrics, and validation strategies for perception and sensor ingestion in maritime environments.

• Build production-quality perception and ingestion pipelines in modern C++ and Python
  , from raw sensor input through onboard runtime deployment.
• Develop and deploy machine learning–based perception models, including optimization for onboard and embedded compute.
• Where appropriate, implement GPU-accelerated components (e.g., CUDA) to meet latency, throughput, and power constraints.
• Own dataset curation, model evaluation, regression testing, and continuous performance improvement.

• Lead and participate directly in field testing, sensor bring-up, and data collection on vessels.
• Design and execute sensor calibration, synchronization, and validation procedures.
• Build robust logging, replay, and annotation workflows to ensure high-quality training and evaluation data.
• Work hands-on in real operating conditions to validate both sensors and ingestion pipelines.

• Partner closely with Mechanical, Electrical, and Systems Engineering teams to:
• Define camera and sensor requirements
• Influence sensor selection, placement, and mounting
• Ensure ingestion pipelines reflect real hardware constraints and failure modes
• Collaborate with autonomy and controls engineers to define perception outputs that enable safe and reliable navigation decisions.

• Serve as the technical owner of perception and sensor ingestion within the autonomy organization.
• Establish engineering best practices, coding standards, and architectural patterns.
• Mentor and help hire future perception engineers as the team grows.

• 7+ years of experience building perception systems for robotics, autonomous vehicles, or complex real-world systems.
• Deep expertise in computer vision and perception, including detection, tracking, and sensor fusion.
• Strong proficiency in modern C++ and Python for production systems.
• Experience designing sensor data ingestion pipelines for high-throughput, time-sensitive systems.
• Demonstrated ability to work across software, hardware, and systems boundaries.
• Comfortable working hands-on with sensors,…