AI working practice
Give the system a problem.
My use of AI developed from conversational assistance into a working environment for research, evidence analysis, software development, workflow design and controlled autonomous execution.
The objective is never to remove human judgement. It is to make judgement more informed, execution more capable and the boundaries of autonomy explicit.
Interactive reasoning flow
Choose a problem, then inspect the decisions behind the output.
The visible result is the last step. The quality of the work depends on how I frame the problem, constrain authority and verify execution.
Selected problem / 2026
CareerAgent
A conventional CV could not preserve the evidence, context and authority boundaries needed to represent a complex career accurately.
Operational local systemProblem
A conventional CV could not preserve the evidence, context and authority boundaries needed to represent a complex career accurately.
My judgement remains central
Different systems. Different roles.
I assign tools to parts of a workflow, not to a logo wall.
The method is orchestration: interpretation, alternative reasoning, local execution, version control and browser reality checks.
Assigned role
Local execution
Works with real files, code, browsers and tests to implement accepted specifications.
Quality control
Generate. Challenge. Verify. Test. Accept or reject.
A polished answer is still rejected when the framing, evidence or authority boundary is wrong. AI extends my working capability; it does not replace my responsibility for the result.
Only evidence-backed, sanitized descriptions appear here. Private prompts, source archives, credentials, counterpart data and internal operating records remain offline.