I'm Christian Macedo — a change and transformation leader with over twenty years designing operating models and delivering complex programmes across regulated industries. Ditching Peter is where I've written down the pattern I keep meeting.
Over two decades my work has taken me across banking (HSBC, Lloyds, Standard Bank), travel, healthcare, and the public sector (NHS Test & Trace, Ofgem) — designing operating models and delivering complex programmes in heavily regulated environments. Different industries, different sizes, the same three questions left unanswered: who decides, who delivers, and who actually has what the role requires.
I think in terms of capabilities — what a role, a team, or a system is accountable for, what it's permitted to do, and what it needs to do it well. It's the same discipline I bring to founding HelixChain, where capability-driven architecture and AI-native product development are applied to gene-editing trials. ARA — Authority, Responsibility, Ability — is that thinking at human scale, stripped back to something anyone can pick up and use on a Monday morning.
This site is deliberately not a book or a course. It's a working artefact: the framework as I actually apply it, kept short enough to be useful.
Diagnosing the same dysfunction over and over isn't a framework — it's a habit. Mapping it turns a private instinct into something structured, portable, and testable. Four reasons drove me to write it down.
Across every organisation I worked in, the pain points traced back to the same place: nobody knew who could decide what, who owed which deliverable, or what the role genuinely required. Naming that pattern precisely was the first step to fixing it — and to giving people a shared language for it.
Solving the problem in the room once helps one team once. Mapping it into a repeatable framework means the next team — or a team I'll never meet — can run the same diagnosis themselves. The value is in it being portable, not bespoke.
Once authority, responsibility, and ability are written down as structured data rather than prose, AI can work with them — drafting job specs, auditing role overlaps, and diagnosing where a person or a prompt is failing. The framework was designed to be machine-legible, not just human-readable.
This is as much a personal artefact as a public one. It's the current state of how I reason about organisational design — a document I refine as I learn, and a reference I return to rather than re-deriving the same argument each time.
If any of this resonates, the framework itself is the best place to go next. Pick a single role — your own is fine — and try the ARA self-check at the end.
— Christian Macedo
Founder, HelixChain · BSc (Hons) Computing & Information Systems