Associate AI/ML Engineer

About the Role

We are looking for an Associate AI/ML Engineer to implement and evaluate machine learning models on a federal data and AI program under the direction of senior engineers. The program applies self-supervised representation learning to tabular and temporal administrative data, uses locally hosted language models for structured information extraction, and trains supervised classifiers on learned embeddings. Work is performed inside an access-restricted analytic environment with approved tooling and CPU-only compute.

This is an early-career modeling position for a candidate with a strong quantitative foundation and one to three years of applied experience. You will implement training and evaluation code, run experiments, measure results against reference sets, and be expected to explain the behavior of the models you work on in mathematical terms.

What You Will Do

Model Implementation and Training

• Implement training and evaluation code for tabular transformers (SAINT, TabTransformer) and gradient-boosted classifiers (LightGBM, XGBoost) in PyTorch and scikit-learn.

• Execute pretraining pilots and full training runs under direction, and record results in MLflow.

• Run ablations and hyperparameter sweeps and summarize findings for senior review.

Language Model Extraction

• Operate locally hosted open-weight language models for structured extraction: prompt variants, output schema enforcement, post-processing.

• Build and run the field-level agreement measurement between model output and human-coded reference samples.

• Prepare stratified evaluation samples and maintain the labeled reference set.

Evaluation and Explainability

• Produce discrimination, calibration and subgroup performance summaries for every model iteration and flag anomalies.

• Generate SHAP attributions and verify their stability across retrains.

• Detect leakage across temporal and source splits and report it.

Supporting Data Work

• Maintain feature engineering code and the reconciliation checks that catch upstream data changes before they reach a model (Python, Spark).

Mathematical Foundation

• Required at the level of a strong undergraduate or master's quantitative program. The interview includes technical questions on this material.

• Probability and statistics -Random variables, expectation, variance, common distributions; estimators and their bias and variance; hypothesis tests and confidence intervals

• Linear algebra - Matrix operations, eigendecomposition, projections; the matrix form of attention and of linear models

• Optimization - Gradient descent and its variants, learning rates, regularization; convex versus non-convex objectives

• Evaluation metrics - ROC and precision-recall, calibration, class-imbalance effects; why accuracy is the wrong metric for a rare outcome

Technical Stack

• Tabular representation learning - SAINT, TabTransformer, FT-Transformer; self-supervised pretraining objectives (masked-feature reconstruction, contrastive)

• Temporal and sequence modeling - Transformer encoders over ordered event sequences (BEHRT-style), temporal convolutional networks

• Language models - Open-weight instruction-tuned models (Llama, Mistral, Qwen families) hosted locally; quantized CPU inference via llama.cpp or equivalent

• Supervised learning - LightGBM, XGBoost, CatBoost; scikit-learn

• Deep learning framework - PyTorch

• Explainability - SHAP (TreeSHAP, KernelSHAP), integrated gradients

• Data and MLOps - Spark, Parquet, pandas, NumPy; MLflow for experiment tracking and model versioning

• Environment- Cloud-hosted, access-restricted analytic environments

What We're Looking For

• One to three years of applied machine learning or data science experience; internships, research assistantships and substantive thesis or capstone work count.

• Python with NumPy, pandas, scikit-learn and PyTorch; working SQL.

• Demonstrated understanding of transformer architectures and of at least one gradient-boosted method.

• Exposure to language model inference beyond API prompting — local hosting, output constraints, or evaluation against labeled data.

• Quantitative coursework or project work covering the Mathematical Foundation areas above.

• Bachelor's or master's degree in mathematics, statistics or applied mathematics with data science coursework or concentration is preferred. Computer science, physics or other quantitative degrees are considered where coursework demonstrates the Mathematical Foundation areas above.

Bonus Points

• Coursework or projects in self-supervised learning, sequence modeling, or model interpretability.

• Experience with model quantization or CPU-inference tooling (llama.cpp, ONNX Runtime).

• Exposure to healthcare, claims, or other administrative data.

• Git-based collaboration and code review experience.

• Eligibility for or prior federal Public Trust determination.

Core Competencies

• Implementation Quality - Writes correct, reviewable training and evaluation code.

• Mathematical Grounding - Explains model behavior in terms of the underlying mathematics.

• Evaluation Rigor - Reports calibration and subgroup metrics, not accuracy alone.

• Extraction Measurement - Measures model output against human-coded references field by field.

• Reproducibility - Records every run so results regenerate.

• Technical Growth - Takes review and applies it to the next iteration.

• data integrity, audit requirements, and critical workflows so the platform works correctly for the people and systems that depend on it.

WHY iADEPTIVE

iAdeptive Technologies is an 8(a) small business that delivers modern data, cloud and AI engineering to federal mission programs. We are engineers first. We win work by building systems that hold up under audit and scale under real load — not by selling slideware. Teams here are small, which means a junior engineer works directly with senior people rather than two layers away from them — and what you build gets used.

DETAILS

• Location: Remote (U.S.); Maryland-area candidates preferred for occasional on-site collaboration.

• Employment Type: Full-time, W-2. Eligibility to work in the U.S. required; this role supports federal programs and requires the ability to obtain a Public Trust or higher background determination.

• Education: Bachelor's or master's degree in mathematics, statistics or applied mathematics with data science coursework or concentration is preferred. Computer science, physics or other quantitative degrees are considered where coursework demonstrates the Mathematical Foundation areas above.

• Experience: One to three years applied, including internships and research positions.

• Benefits: Health, dental and vision coverage; 401(k) with company contribution; paid time off and federal holidays; training and certification support.

The posted salary range reflects the floor and target for this role and may vary by work location. Actual offers depend on experience, certifications, and clearance level.

iAdeptive Technologies is an Equal Opportunity Employer. All qualified applicants will receive consideration for employment without regard to race, color, religion, sex, sexual orientation, gender identity, national origin, disability, protected veteran status, or any other characteristic protected by law.

Pay: $64,000.00 - $86,000.00 per year

Benefits:

• 401(k)

• 401(k) matching

• Dental insurance

• Health insurance

• Life insurance

• Paid time off

• Retirement plan

• Vision insurance

Education:

• Bachelor's (Required)

Experience:

• Engineering: 1 year (Required)

Work Location: Hybrid remote in Columbia, MD 21046