Whitepaper
Governing AI Coding Agents: A Baseline-First Operating Model
An operating model for AI coding agents: one versioned baseline repository, rules changed by pull request, third-party skills pinned and license-gated in code, and a health check that can actually fail.
Whitepaper
High-assurance file delivery: gating downloads behind hardware authentication
A signed download link protects the address, not the person who holds it. This paper describes a way to release sensitive files only after the recipient proves they hold a hardware credential — a security key or a PIV or CAC card — and the choices that keep the control real without making the download a chore.
Whitepaper
Name the Role, Not the Vendor: Interface Naming as an Architectural Decision
Most naming debates deserve five minutes. A few are architectural decisions in disguise. Why an interface name should describe its role rather than the vendor, tenant, host or environment behind it — which names must not be changed, and how to migrate one safely.
Whitepaper
Regulated Rails: Compliance-Grade Tokenization on the XRP Ledger
Tokenization is leaving the pilot lab and meeting the examiners. What compliance-grade digital asset infrastructure actually requires — regulated ledger integration, stablecoin treasury rails, key-management architecture, and sovereign ledger infrastructure — and why the XRP Ledger's design maps to institutional requirements.
Whitepaper
From Signal to Decision: Engineering a Sovereign Intelligence Pipeline
Decision-makers don't lack information — they lack the structure that turns the world's signals into their decisions. The anatomy of an intelligence pipeline that fuses, analyzes, and delivers inside your boundary: ingest, fusion, analytics, need-to-know security, and delivery at the cadence leadership actually works in.
Whitepaper
Private AI Behind the Boundary: An Architecture for Sovereign LLM Operations
Why the organizations with the most to gain from large language models are the least able to send their data to one — and the architecture that resolves the dilemma: private model hosting, retrieval-augmented generation over data that never leaves the boundary, and the governance layer that makes the whole thing defensible.
Whitepaper
Silent Failures on the Headless Seam
In one release we fixed three defects on the boundary between a Drupal CMS and its Next.js front end. None of them produced an error. All three were invisible to CI, types, and health checks. This is why that boundary fails quietly, and what we did to make it fail loudly instead.
Whitepaper
Gating Private Files in a Headless CMS
The engineering behind File Gate: how a decoupled Drupal 11 platform withholds a private file until a visitor has earned it — deny at the source, deliver through a short-lived signed route, and never leak the storage path.
Whitepaper
Collapsing a Content Model in Place
How we collapsed twelve Drupal content types into six on a live headless site — in place, with no downtime, and without losing a single node id, URL, revision, or SEO value.
Whitepaper
Composing Navigation in a Headless CMS
How we built self-maintaining hybrid navigation for headless Drupal 11 and Next.js: editors curate the top level, children populate themselves from published content, and every URL stays canonical.