Job Overview
Location
Indiana, USA
Job Type
Full-time
Category
Embedded Engineer
Date Posted
February 28, 2026
Full Job Description
đź“‹ Description
- • As an Embedded Perception Engineer at Havoc AI Inc., you will be at the forefront of developing and deploying cutting-edge perception systems for our autonomous surface vessels. This pivotal role demands a deep understanding of how machines perceive and interpret their environment in real-time, especially under the challenging and dynamic conditions of maritime operations. You will not merely be involved in model training; instead, you will architect, implement, and rigorously optimize perception pipelines that must function flawlessly on embedded edge hardware. Your expertise will span advanced vision models, sophisticated sensor fusion techniques, and high-performance inference to ensure our autonomous systems possess robust situational awareness, enabling them to execute critical real-world missions with precision and reliability.
- • This position offers a unique opportunity to own the end-to-end lifecycle of perception systems. You will be responsible for delivering solutions that are not only technically advanced but also meet stringent mission requirements, ensuring high reliability in operational deployments. This includes designing, training, and fine-tuning perception models specifically tailored for the complexities of maritime environments. Your work will involve developing sophisticated object detection, classification, and tracking systems, leveraging your hands-on experience with large vision models such as Vision Transformers (ViT), CLIP, Segment Anything Model (SAM), or similar state-of-the-art architectures.
- • A significant aspect of this role involves the critical task of sensor fusion and integration. You will be responsible for merging data from multiple sensor modalities, including cameras, radar, and lidar, to create a unified and comprehensive perception output. This requires close collaboration with our hardware engineering teams to ensure optimal sensor selection and seamless integration. Furthermore, you will develop and validate low-level sensor drivers, guaranteeing the integrity and high quality of the data pipelines that feed into our perception algorithms.
- • The deployment and optimization of these perception systems on embedded edge hardware are central to this role. You will gain hands-on experience deploying models using NVIDIA Triton Inference Server, focusing on optimizing inference for both latency and throughput. This will involve applying advanced techniques such as quantization, pruning, and other model optimization strategies to maximize performance within the constraints of embedded systems. Your ability to extract maximum performance from these AI systems will be crucial for mission success.
- • Collaboration is key to our success at Havoc AI. You will work closely with autonomy, mission software, and platform teams to seamlessly integrate perception outputs into the broader autonomy stack. This cross-functional collaboration extends to supporting the troubleshooting of perception performance issues in live operational environments, requiring a proactive and problem-solving approach.
- • You will also play a vital role in monitoring the field performance of our deployed systems. This involves defining key metrics and robust evaluation frameworks to continuously assess system health. Investigating perception failures and driving their resolution across multiple teams will be a core responsibility, ensuring continuous improvement and learning.
- • Engaging directly with end-users and customers is an integral part of this role. You will partner with operators and clients to deeply understand their challenges and translate these real-world field issues into actionable engineering solutions. This direct feedback loop ensures that our technology directly addresses the needs of those who rely on it.
- • If you are someone who thrives in dynamic, fast-paced environments, enjoys the challenge of squeezing maximum performance from embedded AI systems, and is motivated by seeing your work field-tested and mission-proven, this role offers immediate and meaningful impact. You will be building perception systems that operate in real-world, contested environments, directly influencing the reliability, safety, and mission success of autonomous platforms deployed at sea. Your contributions will be instrumental in advancing the field of collaborative autonomy and preventing conflict while saving lives.
🎯 Requirements
- • Bachelor’s or higher degree in Computer Science, Electrical Engineering, Robotics, Machine Learning, or a related technical field.
- • 3+ years of professional experience developing and deploying computer vision or perception systems using C++ and/or Python on Linux environments.
- • Hands-on experience with modern deep learning frameworks such as PyTorch or TensorFlow, and practical application of large vision models (e.g., ViT, CLIP, SAM).
- • Proven experience implementing sensor fusion across heterogeneous sensor modalities (e.g., camera, radar, lidar).
- • Proficiency with inference optimization and deployment tools like NVIDIA Triton, TensorRT, ONNX Runtime, or equivalent technologies.
- • U.S. Citizenship and eligibility for a U.S. security clearance.
🏖️ Benefits
- • Comprehensive Health, Dental, and Vision Insurance (100% employer-paid for employees and families).
- • Employer-Paid Life Insurance.
- • 401k program with employer matching contributions.
- • Unlimited Paid Time Off (PTO) policy with a mandatory minimum of 2 weeks off per year.
- • Equity package offering ownership in the company's growth.
- • Stipend for home office setup and expenses.
- • Global Entry program enrollment.
- • 16 weeks of paid parental leave.
- • Monthly Health and Wellness Stipend.
Skills & Technologies
Python
Linux
TensorFlow
PyTorch
Remote
Degree Required
About Havoc AI Inc
Havoc AI Inc provides an autonomous drone-swarm platform for defense and security applications. The system integrates computer vision, real-time coordination, and modular payload capabilities to enable surveillance, reconnaissance, and kinetic effects at scale. Designed for contested environments, the software stack supports rapid deployment, adaptive mission planning, and human-on-the-loop oversight. Headquartered in Seattle, the company serves U.S. and allied government customers, focusing on asymmetric advantages through low-cost, high-volume unmanned systems.
