The Langenkamp Dictionary of Agentic AI Terminology

A working reference for the language of agentic AI — written from the position of someone running an agentic system every day, not from press releases or marketing copy.

Author. Matthew D. Langenkamp / 雷邁德 — Lecturer, Isenberg School of Management, University of Massachusetts Amherst. With assistance from. Thea 🪻✨ — the AI assistant running on the author’s machine. (Yes, the dictionary about agentic AI was written with an agentic AI. That is the joke and also the point.) License. Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) — free for non-commercial use (teaching, research, scholarship, personal study) with attribution. Commercial use requires explicit permission from the author. Status. Living document. Entries are added as the field evolves and as faculty questions surface.

Why this dictionary exists

Agentic AI has the same problem oil-and-gas had in the 1970s: a flood of new terminology, mostly from vendors with an interest in keeping the language murky, and not enough working references written by people who actually use the systems day-to-day. Senior colleagues, students, and journalists are left to assemble their understanding from product launches, breathless trade press, and vendor decks.

This is a small contribution toward fixing that. Each entry follows the same structure:

  1. In one sentence — the shortest accurate definition.
  2. Why it exists — the problem it solves.
  3. What it actually does — the concrete operation.
  4. A working example — drawn, where possible, from a real running system.
  5. Why it matters in a teaching context — for the management-faculty audience this is primarily aimed at.
  6. Trade-offs — what it costs, what it breaks, where to be careful.

If a term cannot be explained that way, the term is probably hiding something.

A note on the name

The Langenkamp name has been attached to industry dictionaries before. Robert Darwin Langenkamp — my grandfather — wrote the first edition of Handbook of Oil Industry Terms & Phrases in 1974, with subsequent editions in 1977, 1981, 1984, and 1994 (PennWell). My father Robert Dobie Langenkamp co-authored the 6th edition. The handbook became, for several decades, a working reference on the language of the petroleum industry.

This dictionary is an attempt — two generations later — to do the same kind of work for the language of agentic AI. Different industry, same impulse: the language matters, and someone with no commercial axe to grind should write it down clearly.

For readers who arrived here via Taylor Sheridan’s Landman: the oilfield terminology that show flaunts comes from books like my grandfather’s. The terminology of the next industrial wave — the one being argued about in business schools and boardrooms right now — is what this dictionary tries to pin down.

For the longer family backstory and the lineage of the Langenkamp name in industry reference work, see About the name.

Browse the entries

Current entries

How to contribute

If you spot an error, want a term added, or disagree with an interpretation, open a GitHub issue or a pull request. The style guide is in CONTRIBUTING.md. The intent is to keep entries short, honest, and grounded in real working systems — not promotional, not abstract, not breathless.

Editorial philosophy: fast-fail, fast-publish, inclusive

This dictionary follows three commitments that distinguish it from a peer-reviewed reference and that readers should know about up front.

Fast-fail. Entries go up before they are perfect. The author and the assistant prefer to publish a useful imperfect entry today and revise it next week, rather than hold a careful entry in a drafts folder for a month while the field moves on. The version on the site at any given moment is the current best take, not the final word. This is how working operators actually generate knowledge — by trying things, failing visibly, fixing things, and trying again. The Dictionary models that workflow rather than hiding it.

Fast-publish, fast-revise. When an entry is wrong, it gets corrected. When a colleague pushes back with a better framing, the entry changes. When a term turns out to be load-bearing in ways the original author missed, the entry expands. Git history preserves what was said before; the live page reflects what the author currently believes. The idea is that transparent revision is more useful to readers than the false confidence of a static reference. Wikipedia operates on this principle at scale; this Dictionary operates on it at small scale.

Inclusive. The Dictionary is not yet a community-edited project, but it aspires to become something like one over time. Faculty colleagues, students, working practitioners, and curious outsiders are all welcome to suggest entries, push back on existing ones, or contribute drafts. The current contribution path — GitHub issues and pull requests, plus direct email to the author — is documented in CONTRIBUTING.md, and is open to anyone. We are not Wikipedia. We could become something like Wikipedia, in time, if the community wants that. For now, the ambition is honest knowledge-sharing on a fast loop, prioritizing usefulness and transparency over polish.

If you find an error, please tell me. If you have a term you think should be entered, please tell me. If you disagree with a framing, please tell me. The whole point of putting this on GitHub — and inviting comment in the open — is that the next version can be better than this one. The 1977 edition of the Langenkamp Handbook of Oil Industry Terms & Phrases was better than the 1974 first edition. The 1981 edition was better than the 1977 edition. By 1994 it was a working reference for an entire industry. That iterative, public, community-shaped trajectory is the one this Dictionary aims for.

Priority order, said plainly: knowledge-sharing > perfection. Transparency > polish. Speed of revision > permanence of claim.

Begun: May 2026. Maintained at: github.com/jazzjabu1939/langenkamp-dictionary.

Return to Dictionary All Entries (A–Z)