Experience: 8+ years of related experience
Security clearance level: TS/ SCI W/ Polygraph
US citizenship required
Responsibilities
Agentic Solution Design & Development: Lead the design and development of AI-driven solutions from conception to deployment, building multi-agent workflows in which coordinated agents and prompts work together to produce outputs supporting reporting and dissemination. This includes prototyping, writing production-quality code, and maintaining deployed AI capabilities.
LLM Integration: Integrate existing foundation models into the RAD architecture via APIs and orchestration frameworks, selecting the right model and approach for each task rather than building models from scratch.
Prompt Engineering & Workflow Automation: Design, test, and refine prompts and agent instructions to reliably automate manual tasks, and decompose complex mission workflows into discrete steps an agent pipeline can execute.
Evaluation & Reliability: Develop robust testing, evaluation, and validation strategies for AI-generated outputs, ensuring accuracy, consistency, and appropriate handling of edge cases before outputs reach analysts and reports.
Collaboration & Communication: Serve as a key technical liaison, collaborating with cross-functional teams including system engineers, software developers, and domain experts. Effectively present and articulate recommended AI approaches, discussing the tradeoffs and implications of different implementations with both technical and non-technical stakeholders.
Continuous Improvement: Stay current with the rapidly evolving LLM and agentic AI landscape (new models, orchestration frameworks, and tooling), continuously identifying practical opportunities to apply emerging capabilities to the system.
Required Technical Skills
Strong Python development skills, with experience writing production-quality, maintainable code
Hands-on experience building applications on top of existing LLMs (e.g., via APIs for commercial or self-hosted models)
Experience designing multi-agent or multi-step AI workflows, including prompt chaining, tool use, and agent orchestration (e.g., LangChain, LlamaIndex, or custom frameworks)
Strong prompt engineering skills and experience evaluating/validating LLM outputs
