Recommended Audiobooks
Kevin Cravedi
Founder & Editorial Director
Kevin Cravedi leads the editorial direction behind Recommended Audiobooks, including its evidence standards, Recommendation DNA methodology, narration-first analysis, and the systems used to turn research into practical listening guidance.
Why the site exists
Books are read. Audiobooks are experienced.
Most book pages can tell you what a story is about. An audiobook listener also needs to know what the performance is like, how much attention the listen asks for, whether the pacing works in audio, what mood it creates, and whether that particular recording is the right fit. Recommended Audiobooks was built to answer those listening-first questions with more structure and transparency.
Editorial direction
What Kevin oversees
The goal is not to produce generic summaries at scale. The goal is to keep the recommendation system useful, evidence-aware, and honest about what is known.
Methodology & standards
Defines the Recommendation DNA framework, listening-fit questions, confidence rules, and the quality bar for public pages.
Evidence & source policy
Directs how source types are weighted, how unsupported claims fail closed, and how stronger evidence can replace weaker assumptions.
Narration-first analysis
Keeps book-level evidence separate from recording-specific evidence so the site does not describe a narrator’s performance without support.
Product, QA & refinement
Reviews exceptions and live output, corrects weak patterns, and continually improves how the system explains an audiobook to listeners.
How a page is built
Evidence first. Explanation second.
Recommended Audiobooks uses structured research, automated checks, and AI-assisted editorial synthesis. The technology helps organize and explain evidence; the editorial methodology determines what is allowed to become a public claim.
Collect
Gather bibliographic, publisher, professional, listener, narrator, award, and other relevant evidence.
Resolve
Separate the work from the recording and verify edition, narrator, and source scope before performance claims are admitted.
Structure
Build a knowledge profile and Recommendation DNA from evidence rather than arbitrary scoring.
Synthesize
Use AI-assisted editorial writing to turn approved evidence into natural, listener-facing explanations.
Validate
Apply deterministic checks, evidence gates, confidence rules, and public-renderer safeguards before display.
Transparency
Human-directed, AI-assisted.
Not every audiobook page is manually written line by line. The site’s scale depends on structured research, automation, and AI-assisted editorial synthesis. The aim is not to hide that process, but to make it accountable: claims are evidence-gated, recording-specific assertions are treated separately, confidence is surfaced, and pages can be revised when stronger evidence becomes available.
Editorial direction by Kevin Cravedi · Founder, Recommended Audiobooks