Walter Ullon

Polly.io  ·  New York, NY  ·  Remote

• Led modernization of Polly's analytics platform, migrating 16 production pipelines from legacy batch ETL to Delta Live Tables streaming with Change Data Capture — cutting end-to-end latency from 90 to 22 minutes, improving data freshness from 2-hour batch windows to 15-minute incremental CDC, and eliminating hundreds of millions of duplicate records across 10B+ daily rows.
• Designed ML platform on Databricks for loan volume forecasting, including feature store, experiment tracking, and champion/challenger workflows for production model governance.
• Built an automated knowledge graph spanning 29 production pipelines — enabling instant blast radius analysis before code changes, AI-agent dependency lookups, and auto-publishing 1,240+ field definitions to Confluence via CI/CD-integrated generation.
• Architected a Delta Sharing platform with dynamic view generation and row-level security, onboarding 3 external financial clients (mortgage servicers and hedge funds) with zero manual DDL and self-service configuration.
• Designed agentic workflows using Claude Code and the Anthropic API for CI/CD documentation validation, automated lineage tracking, and developer workflow acceleration.

EZOPS  ·  New York, NY

• Owned the data science product roadmap end-to-end, arbitrating between engineering capacity, data science priorities, and client-facing demands across multiple concurrent platform tracks.
• Led structured discovery cycles and OKR planning that determined investment sequencing across model monitoring, data quality, and reconciliation automation initiatives.
• Defined success metrics and tracking frameworks for DS product launches, enabling data-driven go/no-go decisions on platform releases.
• Served as primary liaison between sales, client success, and data science engineering, translating client requirements into scoped technical specs and managing scope risk on delivery commitments.

EZOPS  ·  New York, NY

• Built supervised ML models for client record deduplication, anomaly detection, and behavior prediction across financial services data.
• Developed time-series forecasting pipelines and statistical reporting infrastructure used across production reconciliation workflows.
• Implemented experiment tracking, model versioning, and champion/challenger evaluation frameworks using MLflow and custom tooling.
• Built NLP pipelines for document classification and entity extraction on unstructured financial data.
• Designed and maintained data pipelines in PySpark and SQL on Databricks, supporting medallion architecture and Delta Lake migration.

Montclair State University  ·  Montclair, NJ

• Top Graduating Senior in the Department of Mathematics.
• Concentration in Applied Mathematics. Research in Population Dynamics and Epidemiology.
• Published: "Early Warning Signals for Epidemic Extinction" — International Journal of Chaos and Dynamics.