https://avectic.com/engineering Engineering notes from Avectic. Architectural patterns, design choices, and field lessons from building deterministic decision infrastructure for AI-assisted workflows. en-us Tue, 22 Apr 2026 09:00:00 +0000 Tue, 22 Apr 2026 09:00:00 +0000 https://avectic.com/engineering/the-bridge-pattern https://avectic.com/engineering/the-bridge-pattern Tue, 22 Apr 2026 09:00:00 +0000 info@avectic.com (Ryan Kamykowski) Ask a language model the same question twice and you may get two different answers. That property is fine for a chat interface and disqualifying for a regulated decision. This post describes the architectural pattern that separates probabilistic interpretation from deterministic decision, and the design choices that make that separation work in production. https://avectic.com/engineering/encoding-expert-rules https://avectic.com/engineering/encoding-expert-rules Tue, 22 Apr 2026 09:01:00 +0000 info@avectic.com (Ryan Kamykowski) If you are building a deterministic decision engine, the most important design decision you will make is how you represent rules. This post describes the Criterion object pattern we use to encode 120+ healthcare payer policies and 67 NCCI bundling rules, the composition layer on top of it, and the three mistakes most teams make on their first rule packs. https://avectic.com/engineering/deterministic-evaluation-layer https://avectic.com/engineering/deterministic-evaluation-layer Tue, 22 Apr 2026 09:02:00 +0000 info@avectic.com (Ryan Kamykowski) The AI industry is shipping consequential-decision systems built on probabilistic components and papering over the reliability gap with disclaimers, explainability tooling, and human-in-the-loop review. None of these substitutes for the deterministic evaluation layer that regulated decisions actually require. https://avectic.com/engineering/audit-trails-that-actually-help https://avectic.com/engineering/audit-trails-that-actually-help Tue, 22 Apr 2026 09:03:00 +0000 info@avectic.com (Ryan Kamykowski) Most audit trails are designed for the institution that owns the system. A small architectural shift builds them for the person on the receiving end of the decision as well, changing how decision systems behave in dispute, in appeal, and in remediation. https://avectic.com/engineering/healthcare-pa-field-report https://avectic.com/engineering/healthcare-pa-field-report Tue, 22 Apr 2026 09:04:00 +0000 info@avectic.com (Ryan Kamykowski) Eighteen months of building a deterministic decision engine for spine surgery prior authorization produced lessons that do not appear in most AI industry discussions. The rules are harder than the AI. The domain experts know things AI teams underestimate. Determinism is a trust property, not just a technical one.