Sr. AL/ML Engineer

Markon•Annapolis Junction, MD

1d•$120,000 - $200,000•Onsite

About The Position

PLEX Solutions, A Part of Markon is looking for a Sr. AI/ML Engineer to support a program in the Annapolis Junction, MD area. This role involves designing and integrating LLM/agent orchestration workflows, tuning model gateway routing logic, developing RAG pipelines, and implementing MCP-based tool-broker integrations. The engineer will also evaluate, benchmark, and tune models and agent behaviors, build automated evaluation pipelines, and partner with data scientists and mission analysts to translate research into production-grade AI capabilities. Additionally, the role requires implementing AI governance guardrails and researching emerging AI tooling to keep the command's AI capability current and adaptable.

Requirements

Active TS/SCI w/ Polygraph
Bachelor’s degree from an accredited college or university in Computer Science, Data Science, or a related discipline
8+ years of experience in engineering related discipline
Advanced proficiency in Python for AI/ML application development
Demonstrated experience building or integrating LLM/agent-based systems in production
Experience with retrieval-augmented generation (RAG), vector databases, and prompt/context engineering
Ability to manage competing priorities in fast-paced mission environments and identify/escalate operational risks early

Nice To Haves

Experience supporting DoD cyber operations, or related mission environments
Experience with frameworks such as LangGraph, Agno, or the NVIDIA NeMo Agent Toolkit
Experience with model inference/serving platforms (NVIDIA NIM, vLLM, Ollama, LiteLLM)
Experience with MCP (Model Context Protocol) or similar tool-broker/integration standards
Experience with AI evaluation frameworks, model benchmarking, and observability (logs/metrics/traces) for AI workloads
Experience supporting DoD AI/cyber mission environments

Responsibilities

Design and integrate LLM/agent orchestration workflows using agentic frameworks such as LangGraph, Agno, and the NVIDIA NeMo Agent Toolkit.
Tune model gateway routing logic to select, evaluate, and optimize across multiple frontier and open-weight models for cost, performance, and mission fit.
Develop and integrate RAG pipelines, vector databases, and memory services to give agents access to mission knowledge and context.
Implement MCP (Model Context Protocol)-based tool-broker integrations so agents can safely and securely access enterprise tools and data.
Evaluate, benchmark, and tune models and agent behaviors; build automated evaluation pipelines to track quality, safety, and performance over time.
Partner with data scientists and mission analysts to translate research and prototypes into production-grade AI capabilities for the command.
Implement AI governance guardrails — content/safety filters, policy enforcement, and approval gates — for agentic workflows, and document model/agent architectures and evaluation results for reuse.
Research and evaluate emerging LLM, agent, and AI tooling to keep the command's AI capability current and adaptable.