Junior Artificial Intelligence (AI) Machine Learning (ML) Engineer

📋 Description

• • Join PingWind’s fast-growing AI/ML practice and help federal agencies harness the power of data to solve national-security-scale problems. As a Junior AI/ML Engineer you will be embedded in cross-functional squads that turn messy, real-world data into production-grade intelligence—everything from predicting cyber-intrusion patterns to automating supply-chain risk scoring.
• • Own the end-to-end data pipeline for at least one client model per quarter: ingest classified and open-source data feeds, build reproducible cleaning notebooks, version every artifact in Git, and publish curated data sets to our secure S3 data lake so that downstream models can be retrained on demand.
• • Partner with senior ML engineers to prototype, train, and validate classical and deep-learning models (Random Forest, XGBoost, CNN, Transformer) that operate on text, time-series, and imagery. You will run hundreds of experiments in Python, log metrics in MLflow, and present findings in weekly sprint reviews to both technical staff and mission stakeholders.
• • Optimize models for accuracy, latency, and memory footprint; convert Jupyter notebooks into modular Python packages, containerize with Docker, and deploy via RESTful micro-services on AWS EKS or Azure AKS. You will monitor live performance with Prometheus/Grafana and trigger retraining jobs when drift is detected.
• • Engineer features that capture domain nuance—e.g., encode bill-of-materials hierarchies for supply-chain risk, extract temporal graph motifs from network logs, or create multimodal embeddings that fuse satellite imagery with textual reports. Your creativity directly improves ROC-AUC and saves analyst hours.
• • Contribute to our internal MLOps template repository: write unit tests with pytest, set up CI/CD with GitHub Actions, and document every function in Sphinx so that the next junior engineer can onboard in hours, not days. You will also participate in code-review rituals that enforce PEP 8, type hints, and security scanning.
• • Communicate complex results to non-technical decision makers through interactive dashboards (Streamlit, Plotly Dash) and concise slide decks that highlight operational impact—reduced false-positive alerts, shortened investigation time, or dollars saved through predictive maintenance.
• • Stay on the cutting edge: allocate 10 % of your time to self-directed R&D—experiment with diffusion models, graph neural networks, or federated learning—and share learnings in monthly “Lunch & Learn” sessions that are attended by our 50+ technical staff across 37 states.
• • Grow rapidly through PingWind’s mentorship program: you will be paired with a Staff ML Engineer who meets with you weekly to set OKRs, review pull requests, and map out a personalized learning path that could include AWS Machine Learning Specialty certification, security clearances, or graduate coursework.
• • Champion responsible AI: evaluate models for bias, document limitations, and implement explainability techniques (SHAP, LIME) so that end-users trust the system and we remain compliant with federal fairness and privacy mandates.