INDEPENDENT PRODUCT SYSTEM
A human-led way to build with specialist AI agents while keeping evidence, authority, decisions and accountability visible.
THINK-STAGE CONCEPT
Reversible interface concept · August 2026This visual makes the operating-model thesis discussable. It does not represent a working Studio, specialist-agent runtime, startup programme or school deployment.
CONTEXT
As AI performs more of the labour required to create software, a person may receive a finished application without understanding why it exists, what shaped it, how it works or how to care for it. The product question is how to preserve speed without removing meaningful authorship and capability.
MY CONTRIBUTION
I authored and versioned an evidence-gated product-development framework, then began shaping a wider Studio concept in which bounded specialist roles, durable evidence and explicit human authority support both real delivery and learning through the work.
WHAT IS TRUE NOW
Executable Product Development Framework v1.1 exists as a versioned subsystem with stage contracts, reusable records and repository validation. The Studio operating model, specialist-agent runtime, interactive-fiction experience and audience routes remain Think-stage hypotheses.
CONTEXT & PROBLEM
Agentic systems may increasingly research, design, build, test and maintain an application from a short request. That capability is valuable, but it creates a second product question: whether the person remains able to explain, judge and steward what enters the world.
The Studio explores how AI can remove labour without silently inheriting purpose, trade-offs or accountability.
PRODUCT THESIS
The working thesis is that every meaningful journey should produce two outcomes: a useful stage-appropriate artefact and a more capable builder.
Human authorship does not require manually writing every line. It means remaining meaningfully involved in purpose, evidence, values, quality, risk and consequential decisions.
OPERATING SYSTEM
Roles are conceived as bounded contracts with a purpose, permitted work, expected evidence, limits and handoff. Agents may propose and execute; named humans retain transition authority and accountability.
The versioned framework provides the current delivery spine through evidence-gated stages, durable decision records and explicit proceed, revise, park or stop decisions.
ONE FOUNDATION, TWO ROUTES
For startups and first-time founders, the opportunity is speed without blindness: use agents to move quickly while retaining enough product and technical literacy to challenge recommendations, understand risk and steward the result.
For schools and learners aged 16+, potentially including Transition Year, the opportunity is to create real digital products while learning industry roles, evidence, agentic AI and responsibility through practice. These are routes to investigate, not validated audiences or deployments, and the school route would require educator, consent and safeguarding work.
MATURITY & NEXT EVIDENCE
Framework v1.1 can be inspected as a coherent versioned subsystem. No wider Studio runtime, interactive-fiction product, startup programme or school experience has been built.
The strongest credible challenge is that guided authorship could become ceremony around work a capable general assistant can already do. The next discovery must identify which moments genuinely improve judgement and stewardship, and which merely add friction.
EVIDENCE & BOUNDARIES
WHAT I CARRY FORWARD
The next gate is evidence that guided authorship produces better understanding, decisions and stewardship without adding needless ceremony. Until then, the wider Studio remains a thesis around a real, versioned framework.
Explore all selected work