Open to opportunities

Ryan Farmacka

@ryanfarmacka

Message

Senior machine learning engineer building reliable Spark pipelines and CI/CD.

United States

Message

What I'm looking for

I’m looking for a team where I can build production-grade ML/Spark pipelines with strong CI/CD, testing, and observability—reducing manual ops while improving data quality for downstream decisions.

I build production-focused ML workflows end to end—training/inference and eligibility/enrollment style data flows—centered on reliable Spark-based pipelines, Delta-style storage patterns, and cloud deployments. I’ve hardened pipelines with CI/CD, automated testing, schema evolution safeguards, idempotent backfills, and deterministic partitioning to reduce production breakages and improve downstream data quality.

I’m especially proud of making systems observable and dependable: I add structured logging, metrics, drift signals, actionable alerting, and runbooks so teams can diagnose issues fast. From regulated, PII-aware handling and least-privilege access patterns to promotion gates, rollback-friendly job versioning, and quality checks for join keys and records, I focus on secure, measurable outcomes that cut manual operations.

Experience

Work history, roles, and key accomplishments

Current

Senior Machine Learning Engineer

Current

ScienceSoft

Oct 2022 - Present (3 years 8 months)

Designed an end-to-end ML workflow template for consulting engagements, adding dataset validation gates, CI checks, and repeatable training jobs. Improved production reliability and data quality by implementing idempotent backfills and deterministic partitioning, and strengthened observability with structured logs, metrics, and actionable alerts.

Python Pyspark Delta Style Table Patterns Observability Metrics Alerting Idempotent Backfills Data Validation Docker Partitioning

Machine Learning Engineer

Connect for Health CO

Sep 2017 - Oct 2022 (5 years 1 month)

Built and maintained PySpark data pipelines for a healthcare marketplace enrollment platform, transforming eligibility, plan, and subsidy data into consistent Delta-style tables. Reduced enrollment verification failures by handling late-arriving updates and retries, adding reconciliation/validation checks, and implementing PII-aware, least-privilege patterns with improved observability.

Pyspark Delta Style Tables Late Arriving Data Handling Data Validation Docker Observability Key Normalization & Join Validation

Junior Data Engineer / Developer

MasTec Network Solutions

Aug 2015 - Aug 2017 (2 years)

Supported telecom network operations by building Python/SQL ETL jobs to normalize event logs for reporting and operational dashboards. Improved reliability and runtime performance by adding retry/backoff and idempotent loads, optimizing Spark partitioning and join strategies, and improving operational support with runbooks and better logging.

Python SQL Apache Spark ETL Retry Strategies Partitioning & Join Strategy Optimization Docker Linting