Meridiona podcast recording session covering automated systems and service programming topics

Transcript Archive

Podcast Episodes in Text

Every conversation Meridiona has recorded since 2022 is available here as a full written transcript. No audio required - read at your own pace, search for specific terms, and reference technical details without scrubbing through a timeline.

01 Latest

When Automation Breaks: Diagnosing Service Failures Before They Cascade

38 min read

Tobias Fennick and systems engineer Reva Calloway walk through 4 real outage scenarios from distributed service pipelines. The conversation covers how monitoring thresholds get misconfigured, why alert fatigue causes engineers to ignore the right signals, and what a practical incident response checklist actually needs to include - not the theoretical version you find in documentation.

Read full transcript

State Machines in Practice: Beyond the Textbook Diagrams

29 min read

Implementating finite state machines in production systems rarely looks like the clean diagrams in academic papers. This episode examines 3 common points where state logic collapses under real traffic - particularly around concurrent transitions and orphaned states that accumulate silently over weeks.

Read full transcript

API Versioning Decisions That You Will Regret in 18 Months

24 min read

Guest Priyanka Osei-Wusu has maintained 6 public API versions simultaneously for a mid-sized logistics platform. She describes the specific decisions - URI versioning vs header-based, deprecation windows, client communication cadence - that made the difference between a manageable migration and one that required a dedicated team for 9 months.

Read full transcript

Scheduled Jobs: The Part of Your Infrastructure Nobody Monitors Properly

31 min read

Cron jobs and scheduled tasks fail silently more often than most teams realise. Luís Brathwaite covers dead-man's switch patterns, idempotency requirements for retry logic, and the specific logging structure that makes it possible to reconstruct what a job actually did three weeks after a data discrepancy is discovered.

Read full transcript

Topics covered across episodes

Each episode focuses on a specific technical problem rather than broad trends. The conversations are structured around concrete scenarios - a specific system, a specific failure mode, or a specific decision point that engineers actually face.

Guests include working engineers, not primarily consultants or speakers. Most episodes reference specific tools, version numbers, and configuration details that make the content directly applicable rather than illustrative.

1 Service orchestration and inter-process communication patterns in distributed architectures with more than 8 independent components
2 Configuration management across deployment environments - the gap between what your config files say and what actually runs in production
3 Automated testing strategies for systems with external dependencies that cannot be fully mocked without losing test validity
4 Event-driven architectures: consumer group management, offset handling, and what happens when a topic accumulates 72 hours of unprocessed messages

Episode focus areas

System Architecture

Automation Patterns

Fault Tolerance

Observability

API Design