Oliver Ernster
Principal Python engineer and software architect: APIs, integration and platform modernisation; 28 years at the seams of systems and author of the Decision Architecture series.
Entry point to my public work. Most of it is local-first, correctness-first and built to run on real machines rather than sit behind a login. Three deep dives below carry the core of the argument; everything else is two clicks away.
Start here
CV
The two-page version of 28 years across defence, fintech, telecoms and startups. Seeking a permanent senior or principal role.
Crank the Code ↗
The Decision Architecture thesis in long form.
GitHub ↗
github.com/oernster, the full collection of source.
The work, in depth
Three projects, four questions each: what was hard, what was chosen, what it costs and what proved it.
Decision Architecture
The four-book thesis and the two instruments that make it executable: Fulcrum and LatencyLab.
The hard problem
Organisational theory cannot be run: claims arrive in prose and reorgs get tested on real people.
The decision
Make it executable. Fulcrum scores structural health 0 to 100 with no randomness: same model, same number, every time.
The trade-off accepted
The units are abstract, not calibrated to your KPIs. And the model never says what to decide, only where decisions can live.
The evidence
Typical archetypes collapse 83.4 to 12.0 as they grow; well-designed ones hold 98.7 to 64.9. Falsifiable by design.
The hard problem
Publishers now ship video, audio, articles and releases; RSS models one form well and push alternatives monetise attention.
The decision
A protocol, not a platform: JSON, pull-only, a semantic superset of RSS 2.0 and Atom, calm enforced at the wire level.
The trade-off accepted
No push and a 300-second poll floor: real-time delivery is impossible, on purpose. A community spec, not an IETF work product.
The evidence
A conformance suite covers every normative statement; Meridian proves the rules against real feeds with 100% branch coverage.
Applications
Shipped applications, led by PigeonPost (email, calendar, contacts) and ClearBudget (solvency forecasting).
The hard problem
Webmail wants engagement, Outlook wants a subscription and Thunderbird's priorities sit in someone else's queue; in budgeting, a monthly total can look safe while the month itself is structurally unsafe.
The decision
Local-first, structurally: PigeonPost keeps mail with your provider (no account, no relay, no telemetry); ClearBudget projects the month day by day instead of reporting where the money went.
The trade-off accepted
The local cache is not encrypted at rest; the site says so and points to OS disk encryption. Openly not for webmail-only users.
The evidence
Installers for Windows, macOS and Linux behind a 100% coverage gate on the core logic. Both are the author's own daily tools.
Everything else
The rest of the catalogue, reachable in two clicks.