AI/ML Platform Engineer

Johns Hopkins Applied Physics Laboratory•Laurel, MD

11h

About The Position

We are seeking a highly motivated AI/ML platform engineer who will contribute to solving cutting edge challenges that didn’t exist until recently — integrating agents with enterprise tools, fine-tuning models for regulated use cases, and building the operational discipline that keeps AI systems reliable as the underlying technology changes monthly. You will be joining a team of engineers and scientists who are at the forefront of APL's mission to provide innovative solutions to critical challenges in the area of missile defense. You will work at the intersection of software engineering, data science, and machine learning, to turn experimental models into scalable, operation-ready solutions. You will design and build multi-agent and single-agent systems using modern open-source frameworks (e.g.: LangChain/LangGraph, CrewAI, AutoGen, and their successors) that automate and augment work across our enterprise and developer tools. You will fine-tune LLMs for specific applications using modern techniques (e.g.: LoRA). Prepare training data, run training jobs on shared infrastructure, evaluate results against task-specific benchmarks, and hand off to the ML team's serving platform. You will research and prototype ML algorithms for specific production tasks, measure their performance and limitations rigorously, and produce clear assessments of their robustness and the confidence we can place in them. You will interact with various data types, formats, and structures in a clear and appropriate manner for algorithm training and testing and perform any data cleaning, normalization, or manipulation as needed. You will create effective visualizations and comprehensive documentation to explain complex topics to a variety of audiences.

Requirements

Possess a Bachelor’s degree in Math, Computer Science, Electrical Engineering or a related field.
Have 2+ years of experience in software engineering, machine learning, or relevant fields.
Have experience building production or production-adjacent LLM-powered or agent-based systems.
Have strong proficiency in Python, including the scientific/ML stack (PyTorch or similar, NumPy, pandas).
Have hands-on experience with LangChain and/or LangGraph.
Have the ability to translate mathematical concepts into well-documented and efficient code.
Have work experience with containerization and running containers using Podman or Docker.
Can effectively communicate ideas and results.
Are able to obtain an Interim Secret level security clearance by your start date and can ultimately obtain Secret level clearance. If selected, you will be subject to a government security clearance investigation and must meet the requirements for access to classified information. Eligibility requirements include U.S. citizenship.

Nice To Haves

Have a Master’s or PhD in Engineering, Math, Computer Science, or a related field.
Have 4+ years of experience in designing and implementing AI/ML algorithms for a variety of datasets.
Are competent in a wide variety of programming languages, including MATLAB, C++, Python, and Java, on both Linux and Windows platforms.
Have experience using high-performance computing structures like GPU 's and CPU clusters.
Have work experience with statistical modeling, computer vision, autonomy, deep networks, adversarial networks, or optimization.
Have work experience with MLOps and associated tools, like MLflow, DVC, or LakeFS.
Have the ability to obtain a Top Secret or TS/SCI clearance.

Responsibilities

Design and build multi-agent and single-agent systems using modern open-source frameworks (e.g.: LangChain/LangGraph, CrewAI, AutoGen, and their successors) that automate and augment work across our enterprise and developer tools.
Fine-tune LLMs for specific applications using modern techniques (e.g.: LoRA). Prepare training data, run training jobs on shared infrastructure, evaluate results against task-specific benchmarks, and hand off to the ML team's serving platform.
Research and prototype ML algorithms for specific production tasks, measure their performance and limitations rigorously, and produce clear assessments of their robustness and the confidence we can place in them.
Interact with various data types, formats, and structures in a clear and appropriate manner for algorithm training and testing and perform any data cleaning, normalization, or manipulation as needed.
Create effective visualizations and comprehensive documentation to explain complex topics to a variety of audiences.