NODA is transforming how unmanned systems collaborate in complex, mission-critical environments by developing next-generation solutions that enable the autonomous orchestration of heterogeneous unmanned systems across air, sea, land, and space with vital applications in the defense, intelligence, and commercial sectors.
Requirements
- 3+ years of experience in AI/ML applications (LLM orchestration, autonomous decision-making, or planning systems).
- Proficiency in Python and frameworks such as LangChain or equivalent.
- Familiarity with constraint solving, planning algorithms, or symbolic reasoning.
- Familiarity with distributed systems and real-time orchestration.
- Experience with multi-agent coordination frameworks or reinforcement learning for planning.
- Understanding of secure coding and adversarial robustness in AI-driven systems.
- Hands-on experience with robotics/autonomy frameworks (ROS2, Gazebo, navigation stacks)
Responsibilities
- Integrate LLM orchestration frameworks (e.g., LangChain, ReAct-style planners) into NODA’s orchestration stack.
- Collaborate with autonomy teams to develop a reasoning framework for task decomposition, task scheduling, and task allocation.
- Implement human-in-the-loop workflows for operator trust and explainability of AI-generated plans.
- Manage the full lifecycle of AI agents (models, prompts, tools, memory) with safe deployment and rollback practices.
- Work with data engineers to pipeline sensor and mission data into structured AI-ingestible formats.
- Validate agentic behaviors through simulation-in-loop testing prior to live deployment.
- Build and maintain evaluation, monitoring, and logging systems to track agent performance, cost, and reliability.
Other
- U.S. Citizen with the ability to obtain a security clearance
- Excellent problem-solving skills and ability to collaborate across disciplines.
- Systems thinker able to connect AI outputs to real-world autonomy stacks.
- Fast learner with adaptability to evolving AI/ML frameworks.
- Thrives in ambiguous, mission-driven problem spaces.
