Staff Applied ML Engineer, Federal/National Security

Description

The company is an incubation firm that builds and invests in fast-scaling technology businesses across healthcare, cyber, and national security.

The federal team is hiring a Staff Applied ML Engineer to ship production AI systems for government and national security work.

You will cover LLMs, fine-tuning and adaptation, agents, RAG, evaluation, and inference, turning quickly changing AI capabilities into dependable software for real operational settings.

This is hands-on production work, not pure research or infrastructure cut off from users. You will sit with Forward Deployed Engineers (FDEs) and software engineers, and sometimes speak directly with customers to learn their operational problems and work out how models, data, agents, and code can solve them.

The platform must accept interchangeable models, including small specialized ones for edge use, and run across cloud, on-premise, and edge environments.

The hard part is not showing that an LLM can do a task. It is making AI reliable enough for production, measurable enough to catch failures, adaptable enough to improve fast, and practical inside real national security environments.

The team is focused on the last mile of agentic AI: getting models deployed against mission requirements instead of stopping at decision support or prototypes.

Responsibilities

• Build production applications, AI agents, and multi-step agent workflows on top of LLMs.
• Own the pipelines for model evaluation, fine-tuning, adaptation, and experimentation.
• Decide when a problem calls for prompting, RAG, fine-tuning, a specialized model, or a mix of these.
• Design thorough evaluation systems that measure model and agent quality, reliability, tool use, and task completion.
• Build retrieval and context layers over structured and unstructured mission data.
• Optimize models and inference for production use.
• Ship model updates based on live performance and user feedback.
• Work with FDEs and customers to turn mission requirements into production AI capabilities.
• Turn what is built for one deployment into reusable platform capabilities.

Requirements

• A substantial record of building ML or AI systems and deploying them to production.
• Hands-on experience with current LLMs and foundation models.
• Deep expertise in several of these areas: LLM fine-tuning and model adaptation, model evaluation, RAG and retrieval systems, agentic systems and tool use, model serving and inference, embeddings and knowledge retrieval, synthetic data, and guardrails and AI reliability.
• Comfortable moving between ML experimentation and production engineering.
• Able to work independently on ambiguous technical problems.
• Willing to work close to users and watch how AI performs in real operational workflows.

Conditions

Washington, DC or Northern Virginia is preferred.

The role may require regular customer-facing work, including onsite meetings, secure-facility work, or travel, depending on program needs.

Salary range is $180,000-240,000, plus bonus and equity. This is the typical range, based on experience, skills, and other factors.

U.S. citizenship is required, along with an Active Secret or Top-Secret security clearance, because the role supports federal and national security customers.

Applications are checked by third-party hiring platforms for accuracy and fraud, flagged applications may be declined, and data is kept only as long as needed and then deleted.

Benefits

• For full-time roles: a career track with potential for rapid advancement as the firm grows, based on strong performance.
• Employer pays 100% of comprehensive health coverage, including medical, dental, and vision, for you and your family.
• Paid maternity and paternity leave of 14 weeks at normal pay.
• Unlimited PTO, subject to management approval.
• Professional development and continued learning opportunities.
• Optional 401K, FSA, and equity incentives.
• Mental health benefits through an outside provider.
• Cost-effective GLP-1 medication through an outside provider.
• Benefits may differ for each incubation.