Data Engineer – Classical Statistics & Machine Learning

BLN24•Mclean, VA

5d•Hybrid

About The Position

BLN24 is seeking a mid-level Data Engineer to support a large-scale data and analytics platform modernization effort for a federal statistical agency client. This is a hybrid role: data engineering (building and maintaining the pipelines that bring data into the platform) and applied data science (using classical statistics and machine learning to analyze that data once it’s available). The ideal candidate is equally comfortable writing production-grade ingestion and transformation code as they are designing and validating a statistical or ML model. This role works closely with SMEs across multiple program areas to understand source data, build reliable ETL/ingestion pipelines, and apply analytical methods — anomaly detection, statistical modeling, and machine learning — to support operational decision-making.

Requirements

Bachelor’s degree in Data Science, Statistics, Computer Science, Engineering, or related field (or equivalent experience)
3–5 years of experience spanning both data engineering and data science/statistical analysis
Strong proficiency in Python, including experience with data engineering libraries (e.g., pandas, PySpark) and statistical/ML libraries (e.g., scikit-learn, statsmodels)
Hands-on experience building and maintaining ETL/ELT pipelines, including ingestion, transformation, and validation logic
Solid grounding in classical statistical methods (hypothesis testing, regression, distributional analysis) and practical machine learning techniques
Experience working with SQL and relational/distributed data systems
Ability to work within a federal data environment, including familiarity with data sensitivity tiers and access/disclosure constraints
Strong communication skills, with the ability to explain technical/statistical concepts to non-technical stakeholders
Requires U.S. citizenship and ability to obtain a public trust or other clearance/suitability determination typical of federal contractor engagements

Nice To Haves

Prior experience supporting federal statistical agencies or other federal data programs
Familiarity with Databricks or modern lakehouse architectures (Spark, Delta Lake, etc.)
Experience with workflow orchestration tools (e.g., Airflow, Databricks Workflows)
Experience designing anomaly-detection or outlier-detection approaches beyond standard threshold-based methods
Exposure to disclosure avoidance concepts or working with regulated/protected government data
Experience working across multiple coding environments (Python, R, SAS) within the same analytics platform
Background in requirements gathering or systems design for enterprise data platforms

Responsibilities

Design, build, and maintain ETL/ELT pipelines to ingest data from multiple source systems into the platform’s central data store
Develop and maintain data ingestion workflows for both batch and near-real-time sources
Implement data validation, cleaning, and transformation logic to ensure data quality and consistency across pipelines
Work within a modern lakehouse/cloud data architecture, optimizing pipeline performance and reliability
Build and maintain data models and schemas that support downstream analytics and reporting needs
Monitor pipeline health, troubleshoot failures, and implement logging/alerting for data quality issues
Document data lineage, transformation logic, and pipeline architecture for governance and reproducibility
Apply classical statistical methods (hypothesis testing, regression, time-series analysis, distributional comparisons) to identify trends, anomalies, and outliers in operational data
Design and implement benchmarking approaches that compare production data against historical, modeled, or external reference values
Develop and evaluate machine learning models where appropriate, balancing predictive performance with interpretability for non-technical stakeholders
Investigate flagged anomalies by digging into underlying data to identify root causes and contributing factors
Work with SMEs to translate operational questions into analytical approaches, and clearly communicate statistical/ML findings and their limitations
Account for data sensitivity classifications and governance requirements when designing analyses and models
Collaborate with visualization-focused team members to ensure outputs of statistical/ML work are presented clearly to stakeholders