SPEAK AI PRODUCT
FLUENTLY.
The 25 words that come up in interviews, standups, and exec reviews. Each one defined in a sentence or two, the way the book teaches it.
- 25 terms
- A to Z order
- 6 with full guides
- A/B testing
Comparing two versions of a product or experience against a defined metric to see which one wins.
- AI agent
An AI system that can use tools, take actions, and adapt its next step toward a goal instead of returning a single answer.
AI agents for PMs- AI Product Manager
The product manager who decides what AI-powered products get built, how they behave, how quality is measured, and when they ship.
The zero-to-hired roadmap- ATS
Applicant Tracking System. The software employers use to collect, filter, and rank applications before a human reads them.
The AI PM resume guide- Context engineering
Designing everything a model sees at request time: instructions, documents, tool results, and the relevant conversation history.
- Embedding
A numerical representation of text or other content that lets a system compare how similar two pieces of information are.
- Evaluation
A structured check of AI behavior against defined criteria and representative examples. “Evals” are how AI PMs define what good means.
AI evals for PMs- Fine-tuning
Additional training that adapts a pre-trained model using more specific data, changing how the model behaves.
- Generative AI
AI that produces new content: text, images, audio, code, or video.
- Guardrail
A product or system control that constrains unwanted model behavior, from blocked topics to human approval steps.
- Hallucination
An AI-generated statement that is incorrect, fabricated, or unsupported by the evidence the model was given.
- Human in the loop
A workflow where a person reviews, corrects, or approves an AI output or action before it counts.
- Inference
Running a trained model to produce an output from an input. Every inference has a cost and a latency.
- Latency
The time between a request and its response. For AI features, one of the biggest drivers of user trust.
- LLM
Large Language Model. A model trained on huge amounts of text that can read, reason over, and generate language.
- North Star metric
The single metric that best represents the core value a product delivers to its users.
- PRD
Product Requirements Document. The problem, the audience, the proposed behavior, the success criteria, and the constraints, in one place.
- Prompt engineering
Designing instructions, examples, and context that steer a model toward the output you want.
- RAG
Retrieval-Augmented Generation. Retrieving relevant information first, then giving it to the model as context so the answer is grounded in real sources.
RAG for PMs- RICE
A prioritization framework that scores ideas by Reach, Impact, Confidence, and Effort.
- SQL
Structured Query Language. The language for querying relational data, and the fastest way for a PM to answer their own questions.
- STAR
Situation, Task, Action, Result. The structure interviewers expect for behavioral answers.
50 AI PM interview questions- Token
A unit of text a language model processes. It may be a word, part of a word, or punctuation. Tokens drive cost and context limits.
- Trade-off
A decision where improving one property (quality, speed, cost, safety) makes another less favorable. The core of AI product sense.
- Vector database
A system for storing and searching embeddings, so a RAG pipeline can find the most relevant documents fast.
No term matches that search.
Try a shorter word, or clear the search to see all 25.
Adapted from the glossary and skill chapters of The AI Product Manager Blueprint, first edition. Want the full picture? Start with RAG, evals, and agents.
BECOME THE AI PRODUCT MANAGER COMPANIES PAY $200K+ TO HIRE.
One proven roadmap to AI PM skills, standout projects, a stronger resume, and the job search that gets you hired.
