Welcome to Details
I created this page to answer questions about kevinX terminology and technology. -Kevin
It starts with the written Content, Knowledge Base, and AI Data Stack Layer, then covers the two product series (CORE and X), FAQs, technical notes, QA process, how kevinX uses RAG chunking (or not) and metadata, plus AI stack and workflow diagrams.
A few basics before you begin: kevinX is self hosted, so there are no API costs, you control your own security, and no PII is shared. It’s also platform agnostic, so it works with virtually any AI chatbot, agent or system, CMS, or LMS.
ACTION CONTENT
Microlessons, ebooks, and playbooks addressing SMB topics.
KNOWLEDGE BASE
Collection of customer, leadership, marketing, and sales content.
AI DATA STACK LAYER
Both RAG chunked and unchunked, metadata-enriched.
✦ AI DATA STACK LAYER CORE series
This series includes source-level segmentation across 500+ microlessons, 25+eBooks, and playbooks. It creates a single AI stack data layer that is RAG-ready with full metadata.
JSONL format optimized for RAG, featuring chunked-text architecture, embedded metadata, stable IDs, one record per line, and an embed-ready full-text field. Each record also includes a paragraphs array, where each entry may contain one or more paragraphs grouped into a single chunk for analysis. Full metadata included: title, description, tips, categories, and keywords. Perfect for nuanced results.
OMNIBUS is for builders creating a generative AI application and want a content collection with RAG-based, chunked-text architecture & embedded metadata. Ships with full content plus five metadata fields: title, description, tips, categories, and keywords. Chunks and metadata are at the source level. Excellent for returning single microlessons, eBooks, playbooks.
✦ AI DATA STACK LAYER X series
This series streamlines the kevinX knowledge base into a conversational AI format. Without its own RAG layer. It is designed for automated ingestion in a client’s RAG pipeline.
XA is purpose-built for chatbots or agents that need concise SMB expertise. It draws from the complete kevinX knowledge base. Perfect for nuanced responses.
XM is perfect for ask me anything chatbots or agents for quick access concise, action-oriented small core business knowledge. It draws exclusively from the kevinX microlesson knowledge base for quick responses.
✦ FAQ
Question: What are the differences between content, knowledge base, and an AI data stack layer?
Answer: (1.) Content consists of individual microlessons, eBooks, and playbooks. (2.) Knowledge base is a large collection of content organized into a single file. (3.) AI data stack layer Core or X Series is a RAG-chunked or unchunked knowledge base file.
Question: What doesn’t the X Series include chunking?
Answer: The X Series does not include chunking because the client’s RAG pipeline handles the chunking as part of its ingestion process. It does however include metada.
Question: What topics are covered?
Answer: Customer engagement, leadership, marketing, sales.
Question: How is it delivered?
Answer: Via an email link to Dropbox. Download folder/files, READMEs, DATA DICTIONARIEs. and metadata catalog.
✦ TECH NOTES
Every product file ships with companion resources:
DATA DICTIONARY – Describes the structure, content, and meaning of data elements for each product file.
METADATA CATALOG – An XLSX with a description, practical tips, categories, and keywords for every microlesson, eBook, and playbook.
README – Tells you what you have and how to use it.
✦ QA PROCESS
kevinX content is not merely converted into multiple formats. Every deployment file type is systematically validated against the parent source content through automated integrity testing, ensuring accuracy.
Our proprietary validation process verifies word counts, character counts, and file parity to maintain clean, reliable content, knowledge bases, and AI stack data layers.
