The Journal
Essays
Long-form writing on AI, philosophy, psychology, and systems thinking.
-
Python’s Gravity Well: Language Choice Shapes Architecture
Python is present in 92% of data pipeline codebases, creating path dependencies that constrain infrastructure for years. Its gravity well requires strategic, not revolutionary, escape.
-
Data Privacy Engineering Is a Data Engineering Discipline
Implementing tokenization and differential privacy at the pipeline level reduced PII exposure incidents by 89% while adding less than 3% to processing time.
-
The OKR Trap: Why OKRs Break at Scale
OKRs created 312 person-hours of overhead per quarter at a 120-person org. The framework fails when used to create alignment rather than represent it.
-
The architecture of a second brain that actually works
-
Building Systems That Explain Themselves: Self-Documenting Architecture
Self-documenting architectural patterns reduced documentation maintenance by 79% and cut new engineer onboarding from 34 days to 12 days across 7 engineering teams.
-
Message Queue Selection Is a Personality Test for Your Architecture
Message queue selection for 13 organizations revealed that the choice between Kafka, RabbitMQ, SQS, and NATS exposes deeper assumptions about consistency, throughput, and operational philosophy.
-
Designing Data Pipelines for Machine Consumers
AI agents consume more analytical data than humans at 3 of 5 organizations I work with. Machine consumers require fundamentally different quality contracts.
-
The AI Ethics Career Path Does Not Exist Yet
Analysis of 87 AI ethics job postings found 72% required technical skills the role never uses, while only 19% mentioned process design, the skill that matters most.
-
When Best Practices Become Worst Practices
Tracking 12 industry-standard practices across 8 organizations found 5 actively harmed the adopter. Practices are context-dependent tools, not universal rules.
-
The Art of the Escalation: When and How to Ask for Help
Delayed escalations cost an average of 4.2 additional engineering days per incident. The median delay between knowing help was needed and asking for it was 2.3 days.