CoffeeCup.tech — Documentation & Knowledge Architecture for AI
Documentation & Knowledge Architecture for AI — Applied Research and Authoring
Key Elements
| Category | Data |
|---|---|
| Organization | CoffeeCup.tech |
| Period | September 2026 – Present |
| Domain | Technical Documentation, Knowledge Architecture & AI |
| Role | Technical Documentation & AI Expert |
| Work | Applied research, experimentation, methodology, authoring, and knowledge sharing |
| Topics | AI-ready documentation, docs-as-data, knowledge representations, Agentic AI, docs-as-code |
| Emerging approaches | Google Open Knowledge Format (OKF), LangChain OpenWiki |
| Publishing | CoffeeCup.tech website, YouTube, LinkedIn |
Project Overview
Since September 2026, my work on CoffeeCup.tech has increasingly focused on a broader question: how should technical documentation and knowledge architectures evolve when their consumers are no longer exclusively human?
I conduct applied research on how technical knowledge can be structured, governed, represented, and delivered for both human and AI use. This work brings together technical documentation, docs-as-code, knowledge management, AI-ready content, docs-as-data, Agentic AI, and AI-oriented knowledge representations.
My objective is threefold: develop new expertise, test emerging approaches in practice, and share knowledge that can help anticipate the future of technical documentation.
Context
My earlier work on structured data, docs-as-code, documentation agents, and AI-ready knowledge progressively led me to look beyond individual documentation pages and consider documentation as a governed knowledge foundation.
I have been exploring structured data since late 2025, initially through Schema.org, JSON-LD, metadata, and SEO/GEO work on CoffeeCup.tech. This progressively raised broader questions around how documentation knowledge could be described, connected, reused, and consumed by systems other than human readers.
AI introduces another step in this evolution. A human reader, a search engine, a RAG system, and an AI agent do not necessarily need the same representation, metadata, context, or level of structure.
This raises documentation architecture questions that go beyond using AI as a writing tool: how to identify authoritative knowledge, expose useful context and relationships, preserve provenance, govern metadata, avoid duplicated knowledge silos, and derive representations suited to different consumers.
My current work explores these questions through research, practical experimentation, long-form technical writing, and video publications.
Key Deliverables
AI-ready Documentation & Knowledge Architecture
I am developing a documentation approach centered on one principle: reuse and enrich governed knowledge rather than create independently maintained copies of the same documentation.
My own Docusaurus website plays an important role in this work. Like many of the technical documentation portals I build for clients, its source content is written in Markdown and versioned in Git. This makes CoffeeCup.tech a practical sandbox where I can test ideas around metadata, structured content, knowledge relationships, docs-as-data, AI-oriented representations, and automated derivation.
I explore how existing documentation sources, metadata, relationships, and contextual information can be combined to produce representations adapted to different consumers while preserving traceability and maintainability.
This includes questions such as:
- how to make documentation reliable and usable as a knowledge source for AI;
- how to separate knowledge from its human-facing presentation;
- which metadata and relationships should be made explicit for machine consumption;
- how to derive AI-oriented representations from existing authoritative sources;
- how docs-as-code pipelines can support these transformations;
- how to preserve governance and provenance across multiple representations.
I also study emerging approaches such as Google Open Knowledge Format (OKF) and LangChain OpenWiki, with the goal of testing them more concretely through a personal knowledge-base / Personal Brain project. This will provide a practical environment for experimenting with structured knowledge, provenance, relationships, machine-oriented representations, and agent consumption.
Applied Research, Authoring & Knowledge Sharing
A significant part of this work consists of turning research and experimentation into reusable professional knowledge.
I publish motion-design videos on YouTube, both as longer narrated videos and as Shorts, to explain technical documentation and AI concepts visually. These are complemented by long-form reference pages and analytical articles on CoffeeCup.tech, as well as LinkedIn publications that summarize ideas, share experiments, and open discussion with other professionals.
Producing these different formats also gives me an opportunity to deepen my practical use of generative AI in several modes: conversational AI, coding assistance, research and synthesis, content review, image and motion-design workflows, voice generation, and experimentation with tools such as ChatGPT, Claude, and ElevenLabs.
This work serves three complementary goals: learning, sharing, and anticipating. I use it to deepen my understanding of emerging approaches, formalize my own documentation architecture thinking, share methods and experiments with the community, and explore how technical documentation may evolve as AI becomes a new consumer of knowledge.
My Role
I lead this work independently as a Technical Documentation & AI Expert, combining documentation strategy, information and knowledge architecture, technical analysis, experimentation, authoring, and visual communication.
My role is not limited to researching AI technologies. I approach these topics from a documentation perspective: identifying what knowledge users and systems need, how it should be structured and governed, which representations are appropriate, and where AI can provide useful documentation services.
This work reflects my current positioning around docs-as-data: treating documentation not simply as published pages, but as structured, governed, reusable knowledge that can support human readers, search engines, RAG systems, agents, and other AI consumers.
Technologies & Tools
AI & emerging knowledge approaches
Google Open Knowledge Format (OKF), LangChain OpenWiki, Agentic AI, RAG, AI-ready knowledge, AI-oriented knowledge representations, knowledge surfaces.
Generative AI & production
ChatGPT, Claude, ElevenLabs, conversational AI, coding assistants, AI-assisted research and review, motion-design workflows.
Knowledge & structured content
Docs-as-data, structured metadata, Schema.org, JSON-LD, front matter, knowledge relationships, governance, provenance.
Documentation architecture & publishing
Docusaurus, Markdown, docs-as-code, GitHub, Git, Sourcetree, Mermaid, multilingual documentation, SEO/GEO.
Public Project
This is a public research and authoring project.
The resulting reference pages, analytical articles, experiments, motion-design videos, Shorts, and experience feedback are progressively published on this website, on YouTube, and on LinkedIn.
© Florence Venisse, Technical Documentation & AI Expert – First version dated 11/09/2026.