Map the evidence
Input: source notes, labels, gaps, and the question that matters.
Output: an evidence view that separates supported claims from unknowns.
Build an evidence matrixPractical AI workflows for research briefs, evidence checks, and decision records that keep source labels and unresolved questions visible.
This collection is for work where a tidy summary is not enough. Use it to make the question, evidence, options, and missing information easier for a person to inspect.
Start with the task that already has enough context to inspect. Treat the generated output as a working surface, keep uncertainty visible, and use the related guide only when it helps the next human decision.
Use this route when the quality of the next decision depends on whether a reader can still inspect the evidence, uncertainty, and trade-offs behind it.
Completion is saved only in this browser, so you can return to the next unfinished step.
Input: source notes, labels, gaps, and the question that matters.
Output: an evidence view that separates supported claims from unknowns.
Build an evidence matrixInput: the evidence view plus the reader and decision needed.
Output: a concise brief with source labels and open questions.
Write the research briefInput: options, evidence, constraints, and a decision owner.
Output: a decision log or memo that preserves what was chosen and why.
Create the decision logInput: a confirmed choice, supporting evidence, and unresolved risk.
Output: a decision memo with a request that can be reviewed.
Draft the decision memo
Create a decision-facing brief with a source register, claim boundaries, and a visible review trail—before anyone treats a summary as a recommendation.
BoundaryUse when sources are ready to inspect.
Use AI to shape scattered project notes into a one-page decision memo that keeps evidence, options, recommendations, and unresolved questions separate.
BoundaryUse when sources are ready to inspect.
Use AI to turn scattered project notes into an inspectable decision record with options, evidence, owners, and unresolved questions.
BoundaryUse when sources are ready to inspect.
Organize interviews, support notes, and survey comments into reviewable themes while keeping source labels, counter-examples, and unanswered questions visible.
BoundaryUse when sources are ready to inspect.
Construct a claim-by-claim inspection table that keeps source type, direct support, limitations, and verification work separate before a brief or recommendation is written.
BoundaryUse when sources are ready to inspect.
A compact framework for stating the task, context, constraints, and output shape before you ask AI to write.
BoundaryUse when you can name the reader and goal.