Data Engineering · Analytics Engineering · Data Platforms

Building data systems people can trust.

Lead Data Engineer with 10+ years of experience across P&C Insurance and SaaS. I design enterprise data platforms, governed analytics, reusable dbt frameworks, orchestration and CI/CD — and build business context that makes trusted data usable by AI.

SQLPythondbt AWSRedshiftSnowflake AirflowPrefectFivetran TableauCI/CDData Modeling Data QualitySemantic LayerMCP
Featured work

Data platforms & architecture

01 · Flagship · SaaS

Enterprise SaaS Data & Analytics Platform

Designed, built, and maintained an enterprise data and analytics platform from the ground up in AWS Redshift, integrating Salesforce, NetSuite, HubSpot, and product/platform data. Developed analytics-ready data models supporting revenue, customer, product usage, marketing, finance, and executive analytics.

Built the engineering ecosystem around the platform, including dbt transformations, automated data quality and reconciliation, Airflow and Prefect orchestration, CI/CD, Tableau analytics, and governed business metrics.

AWS Redshiftdbt SalesforceNetSuite HubSpotFivetran AirflowPrefect TableauGitHub Actions

Stored Procedures in dbt

Designed a pragmatic migration bridge for executing existing database stored procedures inside a dbt-managed environment, enabling incremental modernization instead of requiring an all-at-once rewrite.

Airflow & Prefect Orchestration

Built parameterized Airflow and Prefect workflows for warehouse and dbt workloads, with configurable execution, failure handling, monitoring, and deployment controls.

CI/CD & Deployment Engineering

Built state-aware dbt deployment workflows and a practical pattern for automated pre- and post-deployment SQL when production changes require operations outside normal dbt models.

Data Trust & Audit Analytics

Used audit dimensions, reconciliation, and Tableau exception dashboards to make data-quality issues visible, trace them to operational records, and support correction at the transactional source rather than hiding problems downstream.

ARR Metric & Revenue Analytics

Engineered traceable ARR logic around messy real-world contract dates, late renewals, overlaps, backdated changes, expirations, and cancellations — replacing opaque manual calculations with governed revenue analytics.

Data Profiling

Built profiling tools and a practical data-archaeology approach for discovering useful, trustworthy data through distributions, null patterns, uniqueness, relationships, staleness, and other structural signals.

02 · Flagship · P&C Insurance

Enterprise Insurance Data Warehouse

Designed, developed, tested, and enhanced an enterprise data warehouse for Property & Casualty insurance, integrating complex operational data into analytics-ready dimensional models. The platform supported insurance reporting, actuarial and analytical data feeds, product performance analysis, executive reporting, and downstream data science.

The work spans data warehouse architecture, policy and insurance-domain modeling, SCD2, ETL/ELT, data quality, advanced SQL, analytics enablement, and BI.

P&C InsuranceData Warehouse Dimensional ModelingSCD2 SnowflakeRedshift SQLTableau

Enterprise Insurance Data Feeds (ERIS)

Rebuilt the Enterprise Rate Indication System (ERIS), replacing a slow legacy SAS process run every six months with an automated daily feed that unifies historical data from multiple transactional systems and stays consistent with company dashboards and other data feeds.

Modeling Data

Curated datasets for modeling insurance rates across California for Auto, Home, and Landlord products.

Data Governance

Implementation of the Atlan data catalog to improve data discovery, documentation, governance, and shared understanding of enterprise data assets.

Levenshtein Distance in Data Analysis & Load

Applied string-similarity techniques to practical data analysis and loading problems where source values could not be matched reliably with exact comparisons.

03 · AI + Governed Data

AI-Ready Business Context & Governed Metrics

Built a business-context layer combining ontology, business glossary, entity relationships, business rules, governed metric definitions, reusable AI Skills, dbt semantic models, MetricFlow, and a custom MCP server. The goal: let AI agents discover and query trusted business data while enforcing validated definitions and query constraints.

Business OntologyBusiness Glossary dbt Semantic LayerMetricFlow MCPAI AgentsPython
Reusable solutions

Engineering projects

Focused tools and frameworks for recurring data-engineering problems.

Data Reliability

Schema Drift Resilience & Impact Analysis

Two complementary tools: dbt macros that keep models running safely when source columns disappear and reappear, plus column-lineage analysis to identify downstream impact before changes become incidents.

dbtSchema Drift Column LineageData Reliability
Data Modeling

dbt_scd2_plus

Open-source dbt custom materialization for advanced SCD2: historical batch loading, incremental processing, out-of-order transactions, mixed Type I/Type II behavior, deduplication, and validation tests.

dbtSCD2Snowflake RedshiftPostgreSQL
Salesforce + History

Salesforce History → SCD2

dbt materialization that reconstructs analytics-friendly historical records from Salesforce field-level history logs, producing complete SCD2 versions with validity windows.

SalesforcedbtSCD2 Historical Data
Analytics Platform Operations

Tableau Cloud Operations

Utilities for Tableau Cloud operations including data-source refreshes, schedule inspection, and identifying less-busy Tableau Bridge time slots for refresh workloads.

Tableau CloudTableau Bridge AutomationOperations
Dimensional Modeling

Reverse Balance Fact Tables

A dbt implementation of reverse-balance fact modeling for point-in-time reporting, including use cases such as inventory, occupancy, and earned premium in P&C insurance.

dbtSnowflake Fact ModelingP&C Insurance
Analytics Engineering

dbt Segmentation Tools

Custom dbt materialization for customer/entity segmentation that can translate a SQL dbt model into Python execution in Snowflake, supporting analytical segmentation workflows.

dbtSnowflake PythonSegmentation
Analysis & applied statistics

Data analytics & machine learning

Selected analytical work spanning statistical data-quality analysis and predictive modeling, including applied machine learning in P&C insurance.

Data Quality · Statistical Analysis

Before Using Anomaly Detection: Is Your Data Suitable?

A practical analysis of when statistical anomaly detection is useful — and when unstable business behavior makes a learned “normal” unreliable. Explores coefficient of variation, seasonality, structural breaks, false positives, and maintainability.

Anomaly DetectionStatistics Data Qualitydbt
P&C Insurance · Predictive Modeling

Water Peril Claims Research with XGBoost & GLM Models

Insurance claims research comparing predictive approaches for water peril, combining machine-learning and statistical modeling techniques to explore claim behavior and risk.

P&C InsuranceClaims XGBoostGLM Predictive Modeling
Auto Insurance · Machine Learning

Auto Insurance Risk Classification & Claim Prediction

Applied machine-learning analysis of auto insurance risk, including classification and claim prediction using insurance policy and loss-related data.

Auto InsuranceRisk Classification Claim PredictionMachine Learning