> Markdown version of [/jobs/ext/4399480-senior-data-scientist](https://www.wearedevelopers.com/jobs/ext/4399480-senior-data-scientist). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md). --- # Senior Data Scientist - **Company:** BlueFlag LLP - **Location:** Pittsburgh, PA, United States - **Experience:** Expert - **Contract:** Permanent contract - **Skills:** A/B Testing, Artificial Intelligence, Amazon Web Services, Data Analysis, Microsoft Azure, Big Data, Cluster Analysis, Code Review, Continuous Integration, Distributed Computing Environment, Github, R (Programming Language), Statistical Hypothesis Testing, Python (Programming Language), Machine Learning, NumPy, Query Optimization, Power BI, OpenAI, Standard Sql, Azure Machine Learning, SAS (Software), SciPy, Search Technologies, Software Deployment, Stata, Tableau (Software), Cloud Platform System, Feature Engineering, Sql Optimization, Boosted Trees, Retrieval-Augmented Generation, Large Language Models, SHAP (Shapley Additive Explanations), Prompt Engineering, Apache Spark, Model Validation, Generative AI, Agentic-AI, Git, Pandas, Data Lakes, Pyspark, Scikit Learn, Information Technology, Xgboost, Machine Learning Operations, Feature Importance, Software Version Control, Databricks - **Published:** October 4, 2026 - **Apply:** https://www.indeed.com/viewjob?jk=23a8cbe3cf716965 ## About the Role * Bachelor's degree in Engineering, Computer Science, Statistics, Mathematics, Systems, Business, or a related scientific or technical discipline, and 15+ years of experience (or commensurate experience) * Proficient in Python (pandas, NumPy, SciPy, scikit-learn) and advanced SQL (window functions, CTEs, query tuning) for data analysis * 2+ years of hands-on work on a leading cloud data platform such as Databricks, Azure, AWS, or GCP * Experience across the end-to-end data science workflow, from finding and assessing datasets to production deployment, including experiment tracking and model management with MLflow or similar * Solid grounding in traditional machine learning: supervised and unsupervised methods, gradient-boosted trees (XGBoost, LightGBM), model selection, cross-validation, and hyperparameter tuning * Strong applied statistics: hypothesis testing, regression, sampling, and experimental design * Experience working with large datasets in a distributed environment (Spark/PySpark) * Sound model evaluation practice: picking the right metrics, handling class imbalance, avoiding leakage, and explaining model behavior (for example SHAP or feature importance) * Working knowledge of large language models (LLMs) and agentic AI workflows, including prompt design and RAG patterns * Version control with Git and collaborative development practices (code review, branching, testing) * Ability to explain technical work to non-technical stakeholders and turn ambiguous requests into defined deliverables * US Citizen: Must be a citizen of the United States * Security Clearance: Must be able to obtain a public trust clearance. Must be eligible to work in the United States. Desired * 5+ years as a data scientist, shipping multiple products that run in operation * Working experience with Databricks in Azure, including Unity Catalog, Delta Lake, Databricks Jobs, and Databricks notebooks/Repos * MLOps experience across the lifecycle, such as: * + Model Serving endpoints for real-time inference + Feature engineering and feature tables governed in Unity Catalog + CI/CD for ML with Azure DevOps or GitHub Actions, and Databricks Asset Bundles + Production monitoring for drift and model quality (for example Lakehouse Monitoring) + Champion/challenger or A/B testing of models in production + Automated retraining and model lineage and auditability * Experience with Azure AI services (Azure OpenAI, Azure Machine Learning) or Mosaic AI (Vector Search, Agent Framework) * Prior experience shipping products that use LLMs or AI agents, including evaluation and guardrails * 2+ years building visual insights (dashboards and reports) that support operational needs, with Power BI, Databricks AI/BI dashboards, or Tableau * Prior experience with R, Stata, or SAS * Experience refactoring R, Stata, or SAS codebases to Python, including validating that results match * Experience with VA or federal healthcare data, and handling PHI/PII under federal privacy and security requirements * Familiarity with federal AI governance and responsible AI practices (bias testing, model documentation, human oversight) * Experience training or mentoring analysts and data scientists * Master's or PhD in a quantitative field ## Description * Consult with internal clients to frame business problems as analytical or ML problems, define success metrics, and set realistic scope * Build data products and workflows that support critical operations, from source data discovery through production deployment * Migrate and modernize client workloads from legacy environments (including R, Stata, and SAS) to Python and Databricks, with measurable gains in runtime, cost, or reliability * Develop, train, and validate ML models (classification, regression, forecasting, clustering, anomaly detection) for complex business problems * Engineer features and build reusable, governed feature pipelines on large datasets with PySpark and SQL * Track experiments, register models, and manage model versions and promotion through MLflow * Deploy models for batch scoring and real-time serving, and automate retraining with scheduled jobs and CI/CD pipelines * Monitor models in production for performance, data drift, and data quality, and set thresholds and alerts that trigger review or retraining * Lead AI adoption with clients, including LLM and agentic use cases such as retrieval-augmented generation (RAG), document summarization, and classification * Evaluate LLM and agent outputs for accuracy, groundedness, and safety before and after release * Document models (purpose, data, assumptions, limitations, evaluation results) so they hold up to governance and responsible AI review * Build reference use cases with tutorials, reference code, and training to drive adoption of cloud tooling for concrete business results * Host office hours and pair with client analysts and data scientists to upskill them * Present findings and model results clearly to technical and non-technical audiences, including leadership