About Adelphi
Adelphi's Vision is to build the biggest AI Deployment company for National Security, which requires talent that is extremely good at problem solving, gets excited about National Security challenges, and thrives in software factories. We're executing a Mission that (1) Identifies the core bottlenecks in Warfighting, Intelligence and the Enterprise, (2) Assigns top talent with backgrounds in systems and machine learning engineering, and (3) Finds repeatability in solution development to target similar offices across the DoW/IC. The business has 10x'd in under two years and has built trust with the senior most leaders at the Department of War and Intelligence Community.
About the Role: The Forward Deployed Senior Machine Learning Engineer position requires a mix of software development, LLM Ops, and secdevops practices, resulting in an exciting, fast-paced engineering role. This role requires the ability to provide solutions for the full LLM stack, from the OS, storage, and network up to the API and transport layer. Experience in the defense or intelligence fields is required. You will own end-to-end delivery of agentic, full-stack systems built on top of frontier models, from first prototype to stable production, embedded alongside our defense and intelligence customers.
Location: This role is based in the DC/Metro area - we follow a hybrid work model with three days per week in office or onsite with customers.
Clearance Requirement: An active U.S. Government clearance is required.
Is this you?
We are an AI-native engineering team. We expect every engineer to leverage LLMs and AI tooling as a core part of how they design, build, ship, and operate agentic systems that turn frontier-model capability into mission outcomes.
You use AI coding tools (Claude Code, Cursor, Copilot) daily and instinctively.
You prompt-engineer through complex architecture and debug sessions, leverage LLMs across the full development lifecycle, maintain a clear-eyed view of AI limitations in high-security contexts, and stay current with emerging models and tooling.
Hands-on with modern agent frameworks and SDKs (LangGraph, OpenAI Agents SDK, Claude Agent SDK, AutoGen, or similar) and agent evaluation / observability tooling.
Familiarity with MCP or similar LLM integration frameworks.
You have a clear-eyed view of AI limitations. You know when to trust AI-generated output and when to verify.
Expectations:
Own end-to-end delivery of agentic, full-stack systems from first prototype to stable production, embedded alongside defense and intelligence customers.
Build and deploy ML services leveraging LLMs, embeddings, RAG, and agent orchestration into production environments, including classified and air-gapped ones.
Work directly with customers to scope problems, sequence delivery, and ship novel AI applications under real-world constraints.
Codify repeatable patterns into reusable tools and building blocks that help the team ship faster.
Bonus Points:
Proficiency in infrastructure management (Docker, Kubernetes, AWS).
Expertise in encryption, authentication, Linux systems administration, DevOps, or SRE.
Production experience launching agentic services and forward-deployed AI applications that drive operational value at the customer site.
Experience as a forward-deployed engineer or tech-led delivery role, embedded with mission customers, scoping problems, sequencing delivery, and shipping novel agentic applications under tight constraints
Experience with federated or privacy-preserving data architectures (such as differential privacy and secure enclaves).
Develop and optimize transformer-based temporal vision models for 3D perception, using large-scale driving data and rigorous experimentation to improve autonomous-vehicle scene understanding.
Lead the development and deployment of production-grade machine learning systems, shaping architecture and engineering standards while advising clients and mentoring engineers.
Lead the deployment and operational reliability of production LLM applications and agentic AI workflows, building secure, observable, and scalable services across Kubernetes, data pipelines, and cloud infrastructure.
Lead technical discovery and development of video-based computer vision systems for retail, building end-to-end ML pipelines and action-recognition models while validating performance and production constraints.
Lead the 0-to-1 development of video-based computer vision systems for retail, owning technical experiments, data workflows, model development and deployment to deliver robust action recognition in live environments.
