Senior Autonomy Software Engineer

Odin Dynamics•Long Beach, CA

About The Position

Odin Dynamics is hiring a Senior Autonomy Software Engineer to serve as the Responsible Engineer for the autonomy and perception stack of its autonomous underwater vehicle platform. This hands-on individual-contributor role owns the autonomy software lifecycle, from requirements and architecture through algorithm design, implementation, model training, validation, simulation, integration, field testing, sea-trial support, production support, failure analysis, and continued improvement after deployment. The role combines autonomy, perception, decision-making, planning, machine learning, classical computer vision, sensor fusion, simulation, testing, and mission-level behavior. The engineer will develop algorithms and production software, evaluate behavior against real-world operating conditions, analyze performance using logs and field data, make technical decisions, mentor colleagues, and remain accountable for the autonomy stack’s performance and reliability. Close collaboration is expected with Avionics, Embedded Software, and Test Engineering on sensing architecture, platform integration, and validation in laboratory, field, and sea-trial environments. Odin Dynamics describes itself as an early-stage defense technology company founded in 2025, building autonomous underwater vehicles. The company emphasizes rigorous, first-principles engineering, test evidence, and product outcomes, with autonomy software being developed from the ground up under real-world compute, latency, memory, power, reliability, and observability constraints.

Requirements

Deep practical experience developing autonomy, perception, planning, or robotics software for real-world systems.
Strong proficiency in at least one systems programming language.
Strong proficiency in a scripting language commonly used for autonomy development, data analysis, model development, testing, and automation.
Experience with machine learning or deep learning for perception, classification, detection, tracking, recognition, or related autonomy tasks.
Experience with classical computer vision, machine vision, state estimation, localization, sensor fusion, or related non-ML perception methods.
Experience with path planning, motion planning, obstacle avoidance, mission planning, behavior trees, state machines, or other autonomous decision-making approaches.
Experience integrating autonomy software into real systems with imperfect sensors, incomplete requirements, constrained compute, noisy data, and difficult operating environments.
Experience training, validating, evaluating, and improving models using structured metrics, test datasets, simulation, replay, and field data.
Experience developing reliable production software, rather than only prototypes, research code, notebooks, or isolated algorithms.
Experience writing maintainable code, conducting code reviews, developing automated tests, using static analysis, profiling performance, and debugging complex system behavior.
Experience building or extending simulation, replay, SIL, HIL/HITL, or scenario-based test environments for autonomy validation.
Ability to reason from first principles about autonomy and perception failures, sensor limitations, edge cases, and system-level behavior.
Ability to work from incomplete information, make defensible technical decisions, and rapidly iterate between design, implementation, testing, and refinement.
Ability to work independently with minimal oversight while collaborating across engineering disciplines.
Willingness to support field testing and sea trials when required.
Approximately five or more years of relevant experience is preferred, but the posting encourages exceptional candidates at any experience level and emphasizes demonstrated technical depth, multidisciplinary problem-solving, and delivery of difficult systems.
Employment may be conditioned on authorization to receive software or technology controlled under U.S. export control laws without sponsorship for an export license.

Nice To Haves

Experience with autonomous vehicles, unmanned systems, UAVs, underwater or maritime systems, drones, robotics, aerospace systems, self-driving vehicles, industrial autonomy, or other real-world autonomous platforms.
Experience taking autonomy, perception, planning, or machine-learning software from early architecture through field-tested deployment.
Experience developing safety-critical, mission-critical, or high-reliability software for systems operating in difficult real-world environments.
Experience working with a range of perception and navigation sensing technologies.
Experience developing autonomy software for GPS-denied, communications-constrained, low-observability, or otherwise challenging operating environments.
Experience with machine-learning model development, training, evaluation, deployment, or optimization using modern frameworks.
Experience building simulation, replay, synthetic-data, scenario-generation, Monte Carlo, SIL, HIL/HITL, or model-validation environments for autonomy development.
Experience debugging autonomy behavior using logs, telemetry, visualization tools, field data, and post-mission analysis.
Experience working closely with electrical, embedded software, mechanical, manufacturing, and test engineers on tightly integrated electromechanical systems.

Responsibilities

Serve as the Responsible Engineer for the autonomy and perception stack.
Develop greenfield autonomy software from initial requirements and architecture through field-tested, production-ready implementation.
Own autonomy behavior across mission planning, path planning, localization, state estimation, sensor fusion, perception, obstacle avoidance, recovery behavior, and fault-aware decision-making.
Design and implement algorithms for object detection, recognition, classification, tracking, mapping, navigation, and autonomous decision-making.
Develop machine-learning-based and classical perception approaches, selecting methods based on mission requirements, data availability, compute constraints, reliability, and test evidence.
Train, validate, evaluate, and improve perception and autonomy models using real, simulated, and replayed data.
Build and maintain workflows for model evaluation, scenario replay, regression testing, performance analysis, and failure investigation.
Write, review, test, profile, and debug production software in appropriate systems and autonomy-development languages.
Develop performance-conscious autonomy software within compute, latency, memory, power, reliability, and observability constraints.
Collaborate with Avionics engineers on sensor selection, sensing architecture, sensor trade studies, and practical perception requirements.
Collaborate with Embedded Software engineers on platform integration, sensor interfaces, data pipelines, timing, logging, diagnostics, and deployment.
Work with Test Engineering to validate autonomy behavior in simulation, laboratory, field, and sea-trial environments.
Develop and maintain autonomy test environments, including HOOTL/SIL, HIL/HITL, scenario-based simulation and replay, and model validation and regression testing.
Analyze vehicle behavior from logs, telemetry, simulation results, field tests, and sea trials; identify root causes and drive corrective improvements.
Define metrics for autonomy performance, perception quality, model reliability, mission success, and failure modes.
Build automated development workflows using modern build systems, continuous integration, static analysis, profiling, model-evaluation pipelines, and automated testing.
Produce concise, functional engineering documentation that supports implementation, testing, operation, and future development.
Use AI-assisted development tools while independently validating their outputs and maintaining a first-principles understanding of the resulting system.
Mentor other engineers and contribute to technical reviews while remaining hands-on.
Support field testing and sea trials when required.