Software Engineer, AI/ML

T2S Solutions is seeking an AI/ML Software Engineer to support its R&D team by developing solutions, performing independent verification and validation, and reporting on model and technique effectiveness. The role emphasizes mathematical and data science approaches to AI/ML and integration of solutions into partner DoD systems.

Responsibilities:

• Collaborate with the team on R&D projects throughout their lifecycle, from initial proposal to final wrap up, particularly in how it can/does leverage AI/ML models and products, with a goal of integration into partner DoD systems
• Work with vendors through CRADAs to provide constructive feedback on their products and solutions, to improve accuracy, safety, or meet other requirements of the Branch
• Work with external research partners in academia and industry to guide their research projects to meet the project’s goals and collaborate in building a solution that can provide meaningful value to the Branch
• Collaborate and communicate clearly with internal and external development teams, using teams, source control, ticketing, and code review tools

Requirements:

• Must be a U.S. Citizen with an active Secret Clearance (REQUIRED)
• Bachelor's degree in a field related to AI and ML (Computer Science, Artificial Intelligence, Mathematics, Statistics, Data Science, et al) and 3-5+ years of directly related experience
• Proficiency in at least one of the following languages and libraries used within the field (A core competency will be reading and understanding code written by others, and ability to adapt to different styles/architectures is expected): Python (PyTorch, TensorFlow, or Scikit-learn), Rust, or Go
• Proficiency in Git for version control
• Experience in developing or applying one or more of the following types of models or related activities: Classical machine learning paradigms such as computer vision and object classification, Large Language Models (LLMs): Understanding the internal structure and how they function / Prompt engineering, debugging, and validation
• Reinforcement learning techniques, Both Agentic AI (using reprompting) and classical agents
• Diffusion Models for both Image modalities as well as other modalities
• Neuro-symbolic solutions that tie the classical or generative models to verifiable logic
• Adversarial AI/AI Security such as model poisoning, prompt injection, mcp exploits, etc
• Explainable AI (XAI)
• Experience in deploying models and applications using them in a containerized environment (Kubernetes/Docker)
• TS/SCI Clearance
• Strong proficiency in the mathematical bedrock of AI/ML
• Experience with real-world deployments is a plus
• Experience with air-gapped, secure networks and running models on them
• Knowledge and experience with scaling down models to fit on austere (low resource) platforms