Work

mypylogger

A zero-dependency Python logging library with sensible defaults, clean JSON output, and 96% test coverage for reliable structured logging.

The Problem

Python’s standard logging module has inconsistent defaults and requires boilerplate configuration to produce structured output. Lambda functions and CLI tools often end up with ad-hoc logging that is difficult to parse, search, or correlate across invocations.

The Approach

A zero-dependency library that wraps Python’s logging module with sensible defaults: structured JSON output, consistent field names, and configurable log levels via environment variable. Extracted from production Lambda code and packaged as a reusable library.

The Outcome

Drop-in structured logging for Python Lambda functions and CLI tools. Consistent JSON output that works with CloudWatch Logs Insights and any log aggregation system without custom parsing rules.

Stack

TechnologyPurpose
Python 3.xImplementation language
Python loggingStandard library wrapper
PyPIDistribution

Repository: mypylogger


AWS Well-Architected Alignment
  • Operational Excellence: Structured JSON output parseable by CloudWatch Logs Insights
  • Security: No dependencies; no external network calls; auditable source
  • Reliability: Zero-dependency design eliminates transitive dependency failures in Lambda
  • Performance Efficiency: Minimal overhead; no I/O beyond the standard logging handler
  • Cost Optimization: Structured output reduces CloudWatch query cost vs. unstructured logs
  • Sustainability: Lightweight; no unnecessary compute