🐾

been building a few things

click any card to read more  ·  data engineering, analytics, BI

Data Engineering

Dark Capacity

Analytics · Open Data

Reverse-engineered CPU fleet sizes for 20+ hyperscaler data centers using public signals. ~1.7% error margin.

PythonSQLAG Grid
open original →
Marketing Analytics

Omnichannel Attribution

Multi-touch · Markov chains

End-to-end pipeline modeling 25K+ marketing events, reconstructing 14-day customer journeys.

PostgreSQLMarkovPython
open original →
Product Analytics · IoT

Twinsip

Hardware + Data

Dual-fluid dispensing sipper — CAD, fluid dynamics, and SQL-based consumption analysis for product-market fit.

CADSQLIoT
open original →
BI Engineering

Last-Mile Exception Classifier

Amazon · Last-mile ops

SQL logic + ETL pipelines classifying missorted packages at Amazon. Reduced exception volume by 8.4%.

RedshiftQuickSightSQL
open original →
AI / ML

AI Skill Recommender

Workforce dev · Deloitte

Collaborative filtering model enabling 10,000+ career transitions. Productionized via CI/CD pipelines.

PythonMLCI/CD
open original →
Data Engineering · ETL

Social Insurance Data Lake

Deloitte

Centralized lake ingesting 10+ sources, 500+ objects. Cut data preparation time by 15%.

ParquetSCD2SQL
open original →