Currently seeking an AI/ML Engineer to join a team building machine learning and generative AI capabilities for defense and national security missions. This is a hands-on, early-career engineering role focused on the practical work of turning data into working models: assembling and curating datasets, training and fine-tuning models, and building the evaluation harnesses that tell us whether a model is actually good enough to field. You will work under the direction of senior engineers and scientists on tasks scoped within larger programs, with increasing ownership of discrete technical tasks as you develop. The work spans classical machine learning, deep learning, reinforcement learning, and the rapidly evolving generative AI toolchain, applied to real sensor data and real mission problems in ISR, EW, and advanced sensing. This is a strong fit for an engineer who wants to build depth in applied ML on problems where the data is messy, the ground truth is hard-won, and the results have to hold up under scrutiny.
What You'll Do - Build, clean, label, and version datasets from real and simulated sources, including handling class imbalance, distribution shift, and sparse or noisy labels
- Train, fine-tune, and benchmark models using standard open-source frameworks, running experiments as assigned and documenting results clearly
- Develop and run evaluation pipelines - task-specific metrics, held-out and adversarial test sets, regression suites - and report findings honestly, including negative results
- Support generative AI efforts through prompt development and iteration, structured output design, retrieval-augmented generation pipelines, and comparative evaluation of open-weight and commercial language models
- Assist with parameter-efficient fine-tuning (LoRA/QLoRA) and supervised fine-tuning of open-weight models for domain-specific tasks
- Implement reinforcement learning experiments in standard environments and instrument them for reproducible measurement
- Maintain reproducible experiment tracking, environment configuration, and code hygiene so that results can be regenerated by others
- Contribute data, figures, and written findings to technical reports, program reviews, and oral presentations
- Collaborate with algorithm engineers and domain experts to understand the sensor data and physical phenomenology behind the datasets you work with
What You'll Bring - Bachelor's in computer science, Engineering, Mathematics, Physics, Statistics, or a related technical discipline and 0+ years of experience (internships, co-ops, research assistantships, and substantial academic project work count)
- Working proficiency in Python and the core scientific stack (NumPy, pandas, scikit-learn, matplotlib)
- Familiarity with at least one deep learning framework, preferably PyTorch
- Exposure to the Hugging Face ecosystem (Transformers, Datasets, Tokenizers) or equivalent open-source model tooling
- Understanding of ML fundamentals: train/validation/test discipline, overfitting and regularization, cross-validation, and the metrics appropriate to classification, regression, detection, and generation tasks
- Comfort with Git-based collaborative development
- Clear written and verbal communication, and a willingness to say when a result doesn't hold up
Ways to Stand Out- Experience fine-tuning open-weight language models using PEFT methods (LoRA, QLoRA) or the TRL/Axolotl/Unsloth toolchain
- Hands-on work with RAG systems: embedding models, vector stores, chunking strategies, and retrieval evaluation
- Familiarity with LLM evaluation practice - LLM-as-judge, rubric-based grading, eval harness design, and the failure modes of each
- Experience with experiment tracking and data versioning tools (Weights & Biases, MLflow, DVC)
- Exposure to reinforcement learning libraries (Gymnasium, Stable-Baselines3, CleanRL) or to preference-based post-training methods (DPO, GRPO)
- Experience with agentic workflows, tool-calling, or the Model Context Protocol (MCP)
- Work with non-natural-image modalities: RF, radar, SAR, audio, time series, or multimodal data
- Containerization (Docker) and experience running jobs on GPU clusters or cloud compute
- Master's degree in a related technical discipline
Total compensation for this role is market competitive. The anticipated salary range for this position based out of Dayton, OH is estimated at $70,000 to $96,000 annually. The actual salary will vary based on applicant's experience, skills, and abilities, geographic location as well as other business and organizational needs. SRC offers competitive benefit options, for more details please visit our website.
- Location Syracuse, New York Fairborn, Ohio
- Employment Type Full-Time Salaried
- Experience Required 0+ Years
- Education Required Bachelors Degree
- Security Clearance Requirement Must meet eligibility requirements
- Travel % 10
