Decision case
Material claims
The propositions the recommendation depends on, separated from assumptions and implications.
Due diligence, technology choices, regulatory questions and operating-model decisions rarely arrive as a clean dataset. They evolve through documents, management assertions, public research, specialist input, meetings and counterparty feedback.
The method keeps the decision case explicit as it changes: which claims matter, what supports them, what remains uncertain, what each recommendation depends on, and what would trigger reassessment.
The software supports the method internally. The professional work product is the evidence-backed judgement and the record needed to revisit it.
The propositions the recommendation depends on, separated from assumptions and implications.
What evidence supports, qualifies or contradicts each consequential claim, with source authority visible.
Recommendation logic tied to claims so upstream changes can reopen advice rather than leave it stale.
Questions, evidence gaps and decision-sensitive assumptions that should be tested or explicitly accepted.
A point-in-time record that makes later updates explainable instead of revisionist.
Official records, client documents, management assertions, interviews, vendor material, public research and analyst assumptions can all matter without being treated as equivalent evidence.
Published research can expose failure modes, comparators or better tests. Client-specific conclusions still depend on the actual client evidence and an explicit transferability judgement.
Models can structure, summarize, challenge and propose. Their output does not establish evidence authority, and model agreement is not independent corroboration.
Market, technology, operational or counterparty questions where the conclusion may need to be reconstructed after the decision.
Recommendations mixing vendor claims, internal readiness, external research, control assumptions and implementation dependencies.
Work where authority, applicability, currentness and unresolved uncertainty matter as much as topical relevance.
Boundary: this is not automated audit/certification, legal/tax/engineering judgement, generic engagement-management software or a promise that AI can replace specialist professional judgement.