Verified.
Every claim is checked against a primary source. Every author and publisher under license. No AI-generated noise. No unverified web scrape. Decision-grade, end to end.
The foundation.
The Knowledge Layer.
The Knowledge Layer.
27 Years of Editorial Curation, Engineered for AI.
Verified. Curated. Structured.
The proprietary knowledge architecture behind every getAbstract product, every Microsoft connector, every custom layer we build.
We've been curating knowledge since 1999 — long before AI changed what that meant. Now the same editorial judgment that served millions of readers is what makes your AI worth trusting.
What makes it a layer.
Three properties that turn knowledge into something AI can actually use.
A library isn’t a layer. A layer is verified, curated, and structured for the system underneath. Ours has been engineered for 27 years.
Curated.
Our editors decide what enters the layer. They read the books, follow the field, drop what doesn’t earn the reader’s time. Curation is the work, not an afterthought.
Structured.
Enriched with metadata, mapped to skills, tagged for retrieval. Compressed for tokens, structured for queries. Ready for your AI to read, your people to scan, your auditors to trace.
Where it comes from.
Three sources of knowledge. One editorial standard.
850+
Licensed publisher knowledge
850+ publishers, rights-cleared. The licensed record of the world’s best thinking — most of it behind paywalls your AI can’t legally touch.
50K+
Verified open knowledge
The 1% of the open web worth feeding enterprise AI. A verification engine, not a crawler — every source rights-checked before it enters the layer.
~80%
Your experts’ tacit knowledge
~80% of what your organization knows was never written down — it lives in people (Nonaka & Takeuchi, 1995). Tacit captures it, expert-verified. Explore Tacit →
Most AI runs on one source. The advantage compounds when all three are architected together — and the third is yours alone.
Our publishers.
See the full publisher list
850+ publishers. Everything licensed, everything rights-cleared.
Your AI reads what we read. From academic presses to business classics, from market intelligence firms to peer-reviewed journals. Every source under license, no exceptions.
Harvard Business Review
Wiley
Springer
Penguin Random House
McKinsey & Company
MIT Press
Oxford University Press
Cambridge University Press
Bloomberg
The Economist
Wharton School Press
Stanford Business Books
And many more
Three knowledge layers.
Explore Knowledge for Learning
Explore Knowledge for Enterprise AI
Explore Knowledge for Productivity
One foundation, three levels of depth.
Each layer extends the same editorial standard into a different application. Pick the one your team needs. Combine them when you’re ready.
The Core Knowledge Layer.
Built for L&D. Powers learning across the whole workforce.
50K+ verified knowledge pieces from 850+ publishers, mapped to 900+ skills. Available in every format your learners want and every system your team already runs.
Summaries50K+ in every format
Skill mapping900+ skills
PowersKnowledge for Learning
The Platinum Knowledge Layer.
Built for AI. Engineered for IT compliance and agent-scale deployment.
Microsoft-certified, IT-hardened, structured for retrieval. Plugs into Copilot, Gemini Enterprise, ChatGPT Enterprise, or your own MCP-based agents.
ConnectorMS-certified · Copilot Studio · MCP
ComplianceSOC 2 · ISO 27001 · GDPR
PowersKnowledge for Enterprise AI
The Custom Knowledge Layer.
Built for your field — and your own experts’ tacit knowledge.
Curated for your domain from licensed and verified sources, then extended with the tacit expertise of your own people — captured and verified by getAbstract through Tacit. The part of the layer competitors can’t license or scrape. Explore Tacit →
CoverageCustom curation · Tacit capture · Expert-verified
RefreshDaily, weekly, or per release
PowersKnowledge for Productivity
How it works.
Five steps. One standard. Every time.
01
Source.
Books from 850+ licensed publishers. 50K+ verified online sources. Primary documents from regulators, research institutions, and authoritative publications.
02
Verify.
Every source is rights-cleared and IP-checked. Every claim is traceable to a primary reference. Our editorial team is the gate. AI is downstream.
03
Curate.
Domain editors decide what enters the layer. They drop the noise, condense the signal, and write the structured summary your AI can actually read.
04
Structure.
Tagged with metadata, mapped to 900+ skills, enriched with context. Compressed for tokens, structured for retrieval, ready for every delivery mode.
05
Deliver.
Through certified connectors, MCP, your LMS or LXP, your AI agent, or our concierge in Microsoft Teams. The layer goes where your work happens.
The knowledge layer your AI is missing.
Verified. Curated. Structured.
Pilot in 14 days. No procurement contract required. We bring the integration.