Skip to content

Senior Data Engineer · Analytics Engineering · Fraud & Revenue Data Platforms

Alex Huggler

I build production-grade data platforms that reduce fraud loss, improve revenue visibility, and turn ambiguous business problems into reliable analytics systems.

View projects Download resume Contact me
GitHub LinkedIn
50k rows
raw to Gold in 0.53s
medallion transform, single process
41/41
dbt tests passing
unique, not_null, relationships, expressions
~110k
events/sec fraud-scored
labeled eval set, ground-truth P/R published
44
unit tests green
pytest across all three projects

Measured on each project's local demo path, synthetic data, seed 42 — methodology in every repo's Results section.

Selected work

Featured projects

All projects

Project 01

repo

Real-Time Fraud Signals Pipeline

Simulates fraud-event detection across streaming telecom-style data, with rules + statistical anomalies surfaced in seconds.

Kafka -> Spark Structured Streaming -> Delta -> dbt -> Streamlit, with watermarked windows, exactly-once writes, and dbt tests on the marts.

~110k/s detector throughput

  • Apache Kafka
  • Spark Structured Streaming
  • Delta Lake
  • dbt
  • Streamlit

What this demonstrates

  • Streaming ingestion + watermarks
  • Data quality / dbt tests
  • Fraud analytics architecture
  • Production-style local dev
View interactive demo

Project 02

repo

Telecom Billing Lakehouse

Builds a single source of truth for revenue, ARPU, and churn signals over high-volume Call Detail Record data.

Airflow -> MinIO/S3 -> Iceberg Bronze/Silver/Gold -> Great Expectations contracts -> dbt marts, with Terraform for the AWS variant.

0.53s raw to Gold, 49,998 rows

  • Apache Airflow
  • Apache Iceberg
  • MinIO / S3
  • Great Expectations
  • dbt

What this demonstrates

  • Medallion lakehouse design
  • Data contracts at the boundary
  • dbt modeling for BI
  • Orchestration + IaC
View interactive demo

Project 03

repo

AI-Assisted SQL Optimizer

Closes the gap between 'a query that runs' and 'a query a senior would ship' for Spark SQL and Snowflake workloads.

sqlglot AST + heuristic findings feed a structured Claude prompt; optional EXPLAIN runner scores rewrites against a seeded benchmark corpus.

0.67 keyword overlap vs ground truth

  • Python
  • Typer
  • Anthropic SDK
  • sqlglot
  • pytest

What this demonstrates

  • Applied LLM tooling
  • SQL parsing + heuristics
  • Benchmark methodology
  • Engineer-in-the-loop UX
View interactive demo

Prefer charts to code? Explore the interactive data visualizations built from these same pipelines.

About

Senior data & analytics engineer

I'm a senior data and analytics engineer with six years at AT&T building production data platforms on Azure and Databricks — PySpark ingestion, dbt analytics models, BI delivery, streaming systems, and data quality. I treat tests, observability, runbooks, and data contracts as part of the deliverable, not as follow-ups.

Recent impact includes identifying or enabling $900K+/yr in fraud-loss reduction and supporting analytics for a $2.2B portfolio of revenue and operational data — work that spans streaming detection, lakehouse design, and BI delivery on AT&T's data platform.

Dallas, TX · open to senior data / analytics engineering roles

Certified DP-203 Azure Data Engineering DP-100 Azure AI Engineering

Current focus

Stack

Streaming & processing

  • Apache Kafka
  • PySpark
  • Delta Lake

Lakehouse & warehousing

  • Apache Iceberg
  • Databricks
  • Snowflake

Orchestration & modeling

  • Apache Airflow
  • Azure Data Factory
  • dbt

Languages & IaC

  • Python
  • SQL
  • Terraform

Let's talk

Interested in data platforms, analytics engineering, or BI automation?

I'm open to senior data engineering, analytics engineering, and data platform consulting conversations.

Contact me View resume GitHub