The whole editorial process is public.
Where the articles come from, how each one is analysed and validated, and what we deliberately don't include.
Only what we have the right to use
Since 31 August 2026, we summarise an article only when that specific article carries an open licence that allows it (CC BY, CC BY-SA or CC0), checked one by one. Articles added before that date stay in the library. Those that came from vendor blogs keep their "Vendor Insight" tag.
The pipeline — 8 stages, every article
AI does the heavy lifting; a human editor maintains the gate.
Source monitoring
RSS feeds from curated sources checked daily at 6:00 AM UTC.
Pre-ingestion filter
Title/summary keyword check for HR/people-development relevance. Trusted sources bypass; unrelated topics skipped.
Licence check, then full text
We ask Crossref which licence each article carries. Only CC BY, CC BY-SA and CC0 pass; anything else stays a title and a link, and its text is never fetched or sent to AI. Paywalled and registration-gated articles are rejected too — we only publish what readers can fully access at the source.
Deduplication & classification
SHA-256 hash dedupe. Rule-based topic classifier assigns labels. Off-topic content removed.
AI analysis
Claude Sonnet 4.6 generates structured analysis: executive summary, key insights, frameworks mentioned, critical assessment of methodology and bias.
Fact validation
Automated cross-check of names, numbers, organizations against the source text. Each article gets passed / warning / failed status.
Editorial review
A human editor reviews every article for relevance and scope — does it fit people development and leadership? We confirm it belongs in our library; we don't grade the publisher's research. No AI-only publishing.
Publication
Approved articles indexed in Meilisearch and published to the library — homepage, topic pages, and the Monday Brief.
What we deliberately don't include: paywalled content, articles without a verified open licence, vendor press releases, market-research promo, AI-fabricated claims, anything that fails fact-validation. See Contact for corrections or takedown requests.
How we select sources
Research relevance
Source regularly publishes evidence-based people-development content
Editorial standards
Established organisation with a verifiable editorial process
Active RSS feed
Publicly accessible and actively maintained feed
Reader access
Articles must be readable without paywall, registration, or member-gate
Publication frequency
Regular, consistent output of new content
Open licence
Each article must carry CC BY, CC BY-SA or CC0, checked one by one
Where new articles come from today
Four open-access, peer-reviewed journals. Being open access is a reason to check an article's licence, never a reason to assume it.
What an article on Peoplense actually is
Each article you read here is our editorial summary of a piece of research published elsewhere — an executive summary, key insights, practical takeaways, and the frameworks the source mentions. The full text, charts, figures, and the author’s voice live at the original publisher. We link to it wherever we hold a verified address for the original.
We are a triage and discovery layer: scan us in a few minutes to find what is worth your hour. When a piece looks essential, go read it where it was first published. That is the deal we make with our readers and with the publishers we cite.

