About

A public, experimental knowledge product for product work in the AI era.

What the Personal AI Operating Site is, what it publishes, and how its public knowledge stays bounded.

A public working surface

This is a Personal AI Operating Site: an AI-readable personal knowledge product and a public working surface for notes, methods, and experiments.

I’m building it as a small field notebook, editorial archive, and index of reusable product work. It is live, but still experimental. Some collections are intentionally small or still represented by previews on the Home page.

The public layer includes:

a knowledge archive and long-form Articles
reusable prompts, templates, and references in the Library
labs that document hypotheses, setups, results, and next iterations
projects presented as public artifacts and connected notes
a quiet Collaboration window for relevant conversations

The perspective behind it

I write from the perspective of a Product Manager working across AI Product Ops, agentic AI workflows, and AI-ready product knowledge.

My domain experience includes B2B SaaS and enterprise purchasing workflows, including SRM and source-to-pay contexts. Public writing stays at the level of reusable product patterns. It does not describe employer systems, customer work, internal roadmaps, or confidential operating details.

The recurring question is practical: how can product knowledge become easier for people and agents to find, review, reuse, and improve?

Who is behind it

I’m Wonbeen Lee, a Product Manager working across Agentic AI, enterprise procurement, and Product Ops. Much of my work involves turning ambiguous requirements into shared product decisions, reusable planning systems, and release-ready outcomes across product, design, engineering, and AI teams.

I’m especially interested in moving AI beyond impressive one-off outputs and into repeatable capabilities with explicit context, review, and human judgment. For current career details, see my LinkedIn profile.

What this site is not

This is not:

a generic portfolio or work-history page
an AI influencer landing page
a prompt dump without context or maintenance
a private KB dump

The site is an operating experiment. Useful material may begin small, change shape, or be retired as the structure becomes clearer.

Public/private boundary

The public/private boundary is part of the product, not a publishing afterthought.

Only public or deliberately generalized material belongs here. Private sources, company and customer details, internal links, raw conversations, drafts, and personal records stay outside the public site and its machine-readable outputs.

AI-readable layer

The same public boundary applies to the AI-readable layer:

/llms.txt is the short entry point for agents.
/llms-full.txt is a broader curated context bundle.
The /md/ Markdown mirror provides public content in a simpler format.
/rss.xml and /sitemap.xml expose published updates and routes.

These files are interfaces to curated public knowledge. They are not exports of the private knowledge base.