Writing
Case studies
StackDX · 2025–2026
Rebuilding a US data platform on Dagster
3.8 TB in, 300M+ records served, 945 assets, 15 states
Founder project · 2021–now
One engineer, 18k products: twochannel.ai
18k+ products, 2k+ brands — maintained by LLM-in-the-loop pipelines
Canlin Energy · 2017–2024
From spreadsheets to ML at Canlin
+$730M reserves revisions · 8,000 wells forecast daily · 50+ hrs/week saved
Articles
Designing an end-to-end Machine Learning Workflow for Gas Well Operations
An example of one specific use case for Machine Learning in the Oil & Gas industry, from data collection to model deployment, using a combination of SQL Server, Snowflake (Snowpark), Python, SQL, Streamlit, Tableau and DataRobot
Implementing a FEAST feature store to ease ML model development workflows
An attempt to leverage the open source FEAST feature store to ease the development of machine learning models.
Implementing Tableau as the Primary Data Platform for a Mid Sized Oil and Gas Company
A case study on how to implement Tableau as the primary data platform for a mid sized oil and gas company, including Tableau Prep for ETL, Tableau Desktop for Visualizations, and Tableau Server for a Data Warehouse and collaborative environment.