Started as a contractor through The Select Group. Converted three months early to full-time.Designs PySpark and SQL data pipelines leveraging Ascend DataFlows for batch processing raw search data. Included cleansing PHI for compliance and security. Developed Looker dashboards to automatically perform search analyses. Configured alerts for automatic monitoring of data drift.Managed technical documentation for both databases using tools like Notion, LucidChart, and Alation.

Perfect 5/5 rated online tutor for data science, python, and statistics. Most commonly, students are pursuing a masters degree in AI or Data Science.Most common tutoring subjects are: EDA, statistical models, feature selection, and implementation/interpretation of modelsReviews: https://www.wyzant.com/Tutors/DionysusTutoring

Works with global clients for DE, ML Engineering, CD4ML / MLDevOps, and DS needs. Roles include:Principal Data Engineer - Blend360- Spearheaded the transition of prediction models into production by establishing testing, self-documentation (MLFlow), and monitoring dashboard (Streamlit) with Databricks jobs triggered by Airflow. Created data processing scripts to clean input prediction data in PySpark.- Implemented Spark with MLFlow on an established model resulting in a forecasting time reduction of 90% (30 minutes to 3 minutes).- Refactored data pipeline from Scala to PySpark for maintainability.Data Science Mentor - Time & Technology Inc- Provided guidance on an academic paper describing novel uses of NLP and graph neural networks with forum data.

Consultant through companies The Select Group and Zempleo for client Duke Energy.Tech Lead on internal application that proposes most cost-effective electrical device upgrades based on forecasted electrical load to ensure grid reliability. Integrated A* algorithm with industry tooling using Python for optimal device selection. Worked alongside subject matter experts coding business requirements.Data Scientist on ten-year electrical load forecast using XGBoost model. Designed software architecture to assert reproducibility and clean deployment. Followed agile methodologies.

Consultant through companies The Select Group and Zempleo for client Duke Energy.Designed, defined, and implemented data science pipelines, code structure, and deployments for dependable, reproducible model results.Using PySpark, automated validation of datasets including machine learning model outputs.Supported team members by suggesting improvements, adjusting/maintaining historic models, refactoring codebase, and implementing coding/model/data standards.

Worked on the data science team to create models for clients. In particular, using CNNs to predict of images: beauty, client product placement, and influence on social engagement. Integrated with development and analytics teams to solve cross departmental issues. Included web development, data visualization software, statistical coding, google analytics, database structuring, and automation using Python.Required staying on top of recent tech development and research in regards to machine learning and, more broadly, AI. Understanding what is currently possible, combined with the knowledge to either hand build or implement from 3rd parties, is used to decide which of the two options is more efficient/valuable.