A transparent workflow for source-aware podcast intelligence.
The methodology is designed to prevent opinion from being presented as fact, stale information from being treated as current, and AI output from being mistaken for evidence.
Five-step research workflow
1. Map the source
Identify themes, entities, factual claims, important quotations, and open questions.
2. Classify claims
Separate factual, predictive, interpretive, experiential, opinion, and disputed claims.
3. Verify independently
Prefer primary documents, official data, peer-reviewed research, and transparent reporting.
4. Check freshness
Compare the recording date with later developments for fast-moving subjects.
5. Synthesize for the reader
Organize the final explanation around the reader's question rather than transcript chronology.
Source hierarchy
Primary first
Official documents, original research, direct data, transcripts, and first-party disclosures.
Independent confirmation
Important claims are corroborated when the issue is consequential or disputed.
Uncertainty stays visible
Conflicting evidence and unresolved questions are labeled instead of hidden.