Christopher J. Bratkovics
Data Scientist and Analytics Engineer
Data Scientist and Analytics Engineer with 7+ years in enterprise analytics. I build predictive models, reliable data pipelines, and reporting tools that help people make better business decisions. My work spans data modeling, source integration, validation, and applied AI, with independent projects in forecasting, retrieval, and LLM applications.
View ExperienceExperience
Seven years building analytics and data infrastructure in enterprise advertising
Senior Data Analyst
Data Science / Analytics Engineering
- Built and own the production Snowflake and dbt pipeline unifying five advertising sources for revenue and delivery reporting in Sigma, with S3 feed integration, deduplication, and historical backfills
- Developed Python churn-risk models and K-means segmentation, delivering risk scores and interpretable customer segments through Snowflake and Sigma to guide retention outreach and growth targeting
- Built regression models for inventory utilization and revenue per unit, using cross-market peer clustering to identify performance gaps and support yield-management decisions
- Implemented a Python fuzzy-matching workflow comparing external advertiser names against 400,000+ internal records, delivering Snowflake ID mappings with confidence tiers and business-user overrides
- Created reusable SQL reconciliation checks with explicit tolerances and record-level diagnostics, producing auditable evidence for data-platform migration
- Developed daily programmatic occupancy components in dbt and co-designed a reporting model separating sales activity from shared inventory capacity
- Delivered a generative AI application for CFO financial communications, separating verified SQL data from generated narrative with numeric validation and editable previews
Business Intelligence Data Analyst
Data Architecture / Data Science
- Automated recurring reporting workflows with Python ETL, saving 20+ hours per week across teams
- Designed fact and dimension tables and KPI definitions to support executive dashboards and business reporting
- Built automated data-quality checks and anomaly-detection workflows to identify issues in reporting data
- Developed predictive-model prototypes to support business analysis and decision-making
Education
Master of Science, Applied Data Science
Bay Path University
June 2025
Bachelor of Science, Computer Science
University of Vermont
December 2018
Independent Technical Projects
Self-directed work in forecasting, retrieval, and LLM applications, with source code on GitHub
Ensemble forecasting models and a draft-analysis application using engineered player features. Gaussian mixture modeling and PCA applied to produce probabilistic player tiers, with predictions exposed via FastAPI.
Ensemble models for points, rebounds, and assists using engineered player features, with time-based evaluation routines and predictions served through FastAPI.
A natural-language-to-SQL application using schema inference and schema-aware prompts, with SQL parsing, result previews, and asynchronous query processing.
A chat application integrating OpenAI and Anthropic models with WebSocket streaming, semantic caching, and provider failover including timeouts and retry handling.
A document-ingestion system with chunking, hybrid keyword and vector retrieval, and reranking, connected to question-answering workflows through FastAPI. Includes a companion retrieval pipeline with a RAGAS evaluation harness.
Technical Skills
Tools and methods I use across analytics engineering, modeling, and applied AI
Core
Data Engineering
Modeling and Analysis
Cloud and Development
Applied AI and Applications
Impact
The scale of the analytics and data work I own day to day