Created
March 23, 2026 20:26
-
-
Save 23maverick23/d884c6daf4c56892cce232b18d64a9d9 to your computer and use it in GitHub Desktop.
SCLS: Jeopardy Proctor Guide
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| <!DOCTYPE html> | |
| <html lang="en"> | |
| <head> | |
| <meta charset="UTF-8"> | |
| <meta name="viewport" content="width=device-width, initial-scale=1.0"> | |
| <title>Proctor Cheat Sheet — NetSuite AI Jeopardy</title> | |
| <style> | |
| body { | |
| font-family: Georgia, serif; | |
| font-size: 14px; | |
| line-height: 1.6; | |
| color: #111; | |
| background: #fff; | |
| max-width: 900px; | |
| margin: 0 auto; | |
| padding: 2rem 2rem 4rem; | |
| } | |
| h1 { | |
| font-size: 22px; | |
| font-weight: bold; | |
| margin-bottom: 4px; | |
| } | |
| .subtitle { | |
| font-size: 13px; | |
| color: #555; | |
| margin-bottom: 2rem; | |
| padding-bottom: 1rem; | |
| border-bottom: 2px solid #111; | |
| } | |
| h2 { | |
| font-size: 16px; | |
| font-weight: bold; | |
| text-transform: uppercase; | |
| letter-spacing: .06em; | |
| margin: 2.5rem 0 1rem; | |
| padding-bottom: 4px; | |
| border-bottom: 1px solid #ccc; | |
| } | |
| h3 { | |
| font-size: 13px; | |
| font-weight: bold; | |
| text-transform: uppercase; | |
| letter-spacing: .05em; | |
| color: #444; | |
| margin: 1.5rem 0 .5rem; | |
| } | |
| table { | |
| width: 100%; | |
| border-collapse: collapse; | |
| margin-bottom: .5rem; | |
| font-size: 13px; | |
| } | |
| th { | |
| text-align: left; | |
| font-size: 11px; | |
| font-weight: bold; | |
| text-transform: uppercase; | |
| letter-spacing: .06em; | |
| color: #555; | |
| padding: 5px 8px; | |
| border-bottom: 1px solid #ddd; | |
| background: #f7f7f7; | |
| } | |
| th.val-col { width: 60px; } | |
| th.dd-col { width: 28px; text-align: center; } | |
| td { | |
| padding: 7px 8px; | |
| vertical-align: top; | |
| border-bottom: 1px solid #eee; | |
| } | |
| td.val { font-weight: bold; color: #333; white-space: nowrap; } | |
| td.dd { text-align: center; color: #b8860b; font-weight: bold; } | |
| td.clue { color: #222; } | |
| td.ans { color: #1a5c44; font-weight: 600; } | |
| tr:last-child td { border-bottom: none; } | |
| tr:hover td { background: #fafafa; } | |
| .fj-box { | |
| border: 1px solid #ccc; | |
| border-radius: 4px; | |
| padding: 1rem 1.25rem; | |
| margin-top: .75rem; | |
| background: #f9f9f9; | |
| } | |
| .fj-label { | |
| font-size: 11px; | |
| font-weight: bold; | |
| text-transform: uppercase; | |
| letter-spacing: .07em; | |
| color: #777; | |
| margin-bottom: 4px; | |
| } | |
| .fj-clue { margin-bottom: .75rem; } | |
| .fj-answer { color: #1a5c44; font-weight: 600; } | |
| .rules-box { | |
| background: #f5f5f5; | |
| border: 1px solid #ddd; | |
| border-radius: 4px; | |
| padding: 1rem 1.25rem; | |
| font-size: 13px; | |
| margin-bottom: 2rem; | |
| } | |
| .rules-box ul { margin: .4rem 0 0 1.25rem; padding: 0; } | |
| .rules-box li { margin-bottom: 4px; } | |
| .dd-legend { | |
| font-size: 12px; | |
| color: #777; | |
| margin-bottom: 1.5rem; | |
| } | |
| .dd-star { color: #b8860b; font-weight: bold; } | |
| @media print { | |
| body { padding: 1rem; } | |
| h2 { page-break-before: auto; } | |
| table { page-break-inside: avoid; } | |
| .rules-box { page-break-after: avoid; } | |
| } | |
| </style> | |
| </head> | |
| <body> | |
| <h1>NetSuite AI Jeopardy — Proctor Cheat Sheet</h1> | |
| <div class="subtitle">For facilitator use only · All clues, answers, daily doubles, and Final Jeopardy · Do not distribute to players</div> | |
| <div class="rules-box"> | |
| <strong>Quick rules reminder</strong> | |
| <ul> | |
| <li>Teams must respond in the form of a question ("What is…")</li> | |
| <li>Facilitator judges whether the intent is correct — exact wording not required</li> | |
| <li>Daily Double (<span class="dd-star">★</span>): teams enter wagers before the clue is read; wager added or deducted from score based on correct/incorrect</li> | |
| <li>Final Jeopardy: teams enter wagers before the question is revealed; 30 seconds to write an answer; proctor awards or deducts wager</li> | |
| <li>Use the +100/−100 buttons on scorecards for manual point adjustments as needed</li> | |
| </ul> | |
| </div> | |
| <p class="dd-legend"><span class="dd-star">★</span> = Daily Double · Green answers are the target responses — accept reasonable variations</p> | |
| <!-- ══════════════════════════════════════ | |
| ROUND 1 | |
| ══════════════════════════════════════ --> | |
| <h2>Round 1 ($200 – $1,000)</h2> | |
| <h3>AI Fundamentals</h3> | |
| <table> | |
| <thead><tr><th class="val-col">Value</th><th class="dd-col">★</th><th>Clue</th><th>Answer</th></tr></thead> | |
| <tbody> | |
| <tr><td class="val">$200</td><td class="dd"></td><td class="clue">The property of an AI system where the same input always produces the same output — like a SuiteFlow approval rule.</td><td class="ans">What is deterministic?</td></tr> | |
| <tr><td class="val">$400</td><td class="dd"></td><td class="clue">LLM-based systems are described as this — context-aware and flexible, but outputs may vary for the same input.</td><td class="ans">What is non-deterministic (or probabilistic)?</td></tr> | |
| <tr><td class="val">$600</td><td class="dd"></td><td class="clue">The unit of text — roughly one word — that determines the cost and speed of every LLM input and output.</td><td class="ans">What is a token?</td></tr> | |
| <tr><td class="val">$800</td><td class="dd"></td><td class="clue">This technique reduces AI hallucination by fetching relevant source data from a knowledge base before the model generates its response.</td><td class="ans">What is RAG — Retrieval-Augmented Generation?</td></tr> | |
| <tr><td class="val">$1,000</td><td class="dd dd-star">★</td><td class="clue">The model's 'working memory' — the maximum amount of text it can read and reason over in a single response.</td><td class="ans">What is a context window?</td></tr> | |
| </tbody> | |
| </table> | |
| <h3>The AI Stack</h3> | |
| <table> | |
| <thead><tr><th class="val-col">Value</th><th class="dd-col">★</th><th>Clue</th><th>Answer</th></tr></thead> | |
| <tbody> | |
| <tr><td class="val">$200</td><td class="dd"></td><td class="clue">The bottom layer of the AI stack — ERP, CRM, and HCM systems — described as the source of truth for all AI reasoning.</td><td class="ans">What is the Data & Systems layer?</td></tr> | |
| <tr><td class="val">$400</td><td class="dd"></td><td class="clue">The AI stack layer containing SuiteScript, connectors, and MCP services — what the model can call to act on real data.</td><td class="ans">What is the Tools & APIs layer?</td></tr> | |
| <tr><td class="val">$600</td><td class="dd"></td><td class="clue">The layer where LLMs, rules engines, and policies live — responsible for generating output.</td><td class="ans">What is the Reasoning & Models layer?</td></tr> | |
| <tr><td class="val">$800</td><td class="dd"></td><td class="clue">The top layer of the AI stack — home to task execution, orchestration, and guardrails — and where agents live.</td><td class="ans">What is the AI Automation layer?</td></tr> | |
| <tr><td class="val">$1,000</td><td class="dd"></td><td class="clue">This is why a company with fragmented data across five disconnected systems has a harder AI problem than one on a unified ERP.</td><td class="ans">What is: agents are only as reliable as the data layer beneath them (no single source of truth)?</td></tr> | |
| </tbody> | |
| </table> | |
| <h3>MCP & Protocols</h3> | |
| <table> | |
| <thead><tr><th class="val-col">Value</th><th class="dd-col">★</th><th>Clue</th><th>Answer</th></tr></thead> | |
| <tbody> | |
| <tr><td class="val">$200</td><td class="dd"></td><td class="clue">MCP stands for this — an open, industry-standard interface that lets AI models connect to external tools consistently.</td><td class="ans">What is Model Context Protocol?</td></tr> | |
| <tr><td class="val">$400</td><td class="dd"></td><td class="clue">Why MCP is described as 'the plug, not the thinking.'</td><td class="ans">What is: MCP provides the connection interface — reasoning and decision-making happen in the model layer, not in MCP itself?</td></tr> | |
| <tr><td class="val">$600</td><td class="dd"></td><td class="clue">The correct answer when a prospect asks whether adopting NetSuite's MCP-based AI Connector will lock them into Oracle's infrastructure.</td><td class="ans">What is no — MCP is an open standard, not proprietary to Oracle or NetSuite?</td></tr> | |
| <tr><td class="val">$800</td><td class="dd"></td><td class="clue">MCP serves as the foundation of this NetSuite capability that enables custom agentic-style workflows today.</td><td class="ans">What is the NetSuite AI Connector?</td></tr> | |
| <tr><td class="val">$1,000</td><td class="dd"></td><td class="clue">This is the honest answer when a customer asks 'Is MCP required to use AI in NetSuite?'</td><td class="ans">What is no — MCP is not required; it enables the AI Connector ecosystem but AI features work without it?</td></tr> | |
| </tbody> | |
| </table> | |
| <h3>Human-in-the-Loop</h3> | |
| <table> | |
| <thead><tr><th class="val-col">Value</th><th class="dd-col">★</th><th>Clue</th><th>Answer</th></tr></thead> | |
| <tbody> | |
| <tr><td class="val">$200</td><td class="dd"></td><td class="clue">The HITL category where AI acts fully autonomously — appropriate for low-risk, high-volume tasks like categorizing invoices.</td><td class="ans">What is Automate?</td></tr> | |
| <tr><td class="val">$400</td><td class="dd"></td><td class="clue">The HITL pattern where AI generates a recommendation or draft, but no action executes until a human explicitly approves it.</td><td class="ans">What is Suggest and Confirm?</td></tr> | |
| <tr><td class="val">$600</td><td class="dd dd-star">★</td><td class="clue">Three examples of actions that must always remain in the 'Always Human' category — AI must never act alone on these.</td><td class="ans">What are employee data changes, GL overrides, and bank detail updates? (accept any two)</td></tr> | |
| <tr><td class="val">$800</td><td class="dd"></td><td class="clue">In a SuiteAgents workflow, this is why the agent asks for confirmation before sending a collections email — even though it retrieved the data autonomously.</td><td class="ans">What is: sending an email is an action tool that changes state, requiring human approval under HITL?</td></tr> | |
| <tr><td class="val">$1,000</td><td class="dd"></td><td class="clue">The most important reason a finance leader accepts AI in their ERP workflow — it comes down to this design principle.</td><td class="ans">What is deterministic controls around probabilistic reasoning?</td></tr> | |
| </tbody> | |
| </table> | |
| <!-- ══════════════════════════════════════ | |
| ROUND 2 | |
| ══════════════════════════════════════ --> | |
| <h2>Round 2 ($400 – $2,000)</h2> | |
| <h3>Agentic Levels</h3> | |
| <table> | |
| <thead><tr><th class="val-col">Value</th><th class="dd-col">★</th><th>Clue</th><th>Answer</th></tr></thead> | |
| <tbody> | |
| <tr><td class="val">$400</td><td class="dd"></td><td class="clue">The AI maturity level covering Q&A and knowledge retrieval — where SuiteAnswers and Ask Oracle operate.</td><td class="ans">What is Answering (Level 1)?</td></tr> | |
| <tr><td class="val">$800</td><td class="dd"></td><td class="clue">The maturity level covering drafting, summarizing, and suggesting — where the GenAI API and Exception Management sit.</td><td class="ans">What is Assisting (Level 2)?</td></tr> | |
| <tr><td class="val">$1,200</td><td class="dd"></td><td class="clue">The maturity level for tool use, writebacks, and transactions — where Ask Oracle and AI Connector operate today.</td><td class="ans">What is Acting (Level 3)?</td></tr> | |
| <tr><td class="val">$1,600</td><td class="dd"></td><td class="clue">The maturity level SuiteAgents targets — orchestrating multi-step workflows across multiple systems.</td><td class="ans">What is Coordinating (Level 4)?</td></tr> | |
| <tr><td class="val">$2,000</td><td class="dd"></td><td class="clue">The highest rung on the AI maturity ladder — where the system learns from outcomes and refines its own behavior — currently described as emerging.</td><td class="ans">What is Optimizing (Level 5)?</td></tr> | |
| </tbody> | |
| </table> | |
| <h3>Workflow Spectrum</h3> | |
| <table> | |
| <thead><tr><th class="val-col">Value</th><th class="dd-col">★</th><th>Clue</th><th>Answer</th></tr></thead> | |
| <tbody> | |
| <tr><td class="val">$400</td><td class="dd"></td><td class="clue">In the transportation analogy, a train on fixed tracks maps to this type of workflow.</td><td class="ans">What is a traditional deterministic workflow?</td></tr> | |
| <tr><td class="val">$800</td><td class="dd"></td><td class="clue">A GPS app that reroutes around traffic but always heads to the same destination — the analogy for this workflow type.</td><td class="ans">What is an AI-enhanced workflow?</td></tr> | |
| <tr><td class="val">$1,200</td><td class="dd"></td><td class="clue">A self-driving car where you only set the goal maps to this — the highest level of the workflow spectrum.</td><td class="ans">What is a fully agentic workflow?</td></tr> | |
| <tr><td class="val">$1,600</td><td class="dd"></td><td class="clue">In an AI-enhanced workflow, this tool still owns the overall sequence even when AI steps are inserted.</td><td class="ans">What is SuiteFlow?</td></tr> | |
| <tr><td class="val">$2,000</td><td class="dd"></td><td class="clue">The single word that best describes what an agent does at runtime that a pre-scripted AI workflow cannot — it makes this.</td><td class="ans">What is a decision (or: dynamically determines the next step)?</td></tr> | |
| </tbody> | |
| </table> | |
| <h3>NetSuite AI Features</h3> | |
| <table> | |
| <thead><tr><th class="val-col">Value</th><th class="dd-col">★</th><th>Clue</th><th>Answer</th></tr></thead> | |
| <tbody> | |
| <tr><td class="val">$400</td><td class="dd"></td><td class="clue">This NetSuite feature lets users ask business questions in plain English and get answers from their own ERP data.</td><td class="ans">What is Ask Oracle?</td></tr> | |
| <tr><td class="val">$800</td><td class="dd"></td><td class="clue">The NetSuite AI capability that automatically surfaces invoice mismatches, duplicate records, and payment anomalies for finance teams.</td><td class="ans">What is AI Exception Management?</td></tr> | |
| <tr><td class="val">$1,200</td><td class="dd"></td><td class="clue">The NetSuite tool that lets developers embed custom generative AI capabilities directly into scripts and workflows.</td><td class="ans">What is the GenAI API?</td></tr> | |
| <tr><td class="val">$1,600</td><td class="dd"></td><td class="clue">This upcoming NetSuite tool will let customers build and configure their own AI agents — no deep coding required.</td><td class="ans">What is Agent Studio?</td></tr> | |
| <tr><td class="val">$2,000</td><td class="dd"></td><td class="clue">The NetSuite capability coming in NetSuite Next that orchestrates multi-step agentic workflows — the subject of this entire training.</td><td class="ans">What is SuiteAgents?</td></tr> | |
| </tbody> | |
| </table> | |
| <h3>Positioning & Objections</h3> | |
| <table> | |
| <thead><tr><th class="val-col">Value</th><th class="dd-col">★</th><th>Clue</th><th>Answer</th></tr></thead> | |
| <tbody> | |
| <tr><td class="val">$400</td><td class="dd"></td><td class="clue">The first question to ask any prospect who opens with 'We want agentic AI' — before demoing anything.</td><td class="ans">What is: 'What specific problem are you trying to solve?'</td></tr> | |
| <tr><td class="val">$800</td><td class="dd"></td><td class="clue">Why a prospect's concern about vendor lock-in is directly answered by confirming MCP's status as this.</td><td class="ans">What is an open, industry-standard protocol (not proprietary to Oracle)?</td></tr> | |
| <tr><td class="val">$1,200</td><td class="dd"></td><td class="clue">The positioning phrase that reassures a CFO or auditor that NetSuite AI doesn't sacrifice control for flexibility.</td><td class="ans">What is: deterministic controls around probabilistic reasoning?</td></tr> | |
| <tr><td class="val">$1,600</td><td class="dd"></td><td class="clue">When a skeptical buyer asks 'How do you stop the AI from making things up in my financial data?', this is the technical answer.</td><td class="ans">What is RAG — the AI retrieves structured data from NetSuite before generating, grounding the output in real records?</td></tr> | |
| <tr><td class="val">$2,000</td><td class="dd dd-star">★</td><td class="clue">The career-limiting risk of overselling AI autonomy before an implementation — what happens post-go-live.</td><td class="ans">What is: the customer loses trust when the system requires more human involvement than was promised?</td></tr> | |
| </tbody> | |
| </table> | |
| <h3>Structured vs Unstructured</h3> | |
| <table> | |
| <thead><tr><th class="val-col">Value</th><th class="dd-col">★</th><th>Clue</th><th>Answer</th></tr></thead> | |
| <tbody> | |
| <tr><td class="val">$400</td><td class="dd"></td><td class="clue">Why ERP data — rows, schemas, defined fields — is particularly well-suited for AI compared to emails and PDFs.</td><td class="ans">What is: structured data is deterministic, grounded, and auditable — outputs can be traced to source records?</td></tr> | |
| <tr><td class="val">$800</td><td class="dd"></td><td class="clue">What AI must do with unstructured inputs like emails or PDFs that it does not need to do with structured ERP data.</td><td class="ans">What is interpret (or: infer meaning using probabilistic reasoning)?</td></tr> | |
| <tr><td class="val">$1,200</td><td class="dd"></td><td class="clue">The type of data where 'the same document can yield different outputs' — a key governance risk to explain to customers.</td><td class="ans">What is unstructured data?</td></tr> | |
| <tr><td class="val">$1,600</td><td class="dd"></td><td class="clue">This is why an AI agent working from a NetSuite invoice record produces a more auditable output than one working from a scanned PDF of the same invoice.</td><td class="ans">What is: the ERP record is structured and traceable; the PDF requires inference to interpret?</td></tr> | |
| <tr><td class="val">$2,000</td><td class="dd"></td><td class="clue">The term for grounding AI responses in real, retrieved data — the technique that makes structured ERP data NetSuite's biggest AI advantage.</td><td class="ans">What is RAG — Retrieval-Augmented Generation?</td></tr> | |
| </tbody> | |
| </table> | |
| <h3>SuiteAgents</h3> | |
| <table> | |
| <thead><tr><th class="val-col">Value</th><th class="dd-col">★</th><th>Clue</th><th>Answer</th></tr></thead> | |
| <tbody> | |
| <tr><td class="val">$400</td><td class="dd"></td><td class="clue">In a SuiteAgents cashflow workflow, these tools retrieve AP balances, AR data, and forecasts — without requiring human confirmation.</td><td class="ans">What are read tools?</td></tr> | |
| <tr><td class="val">$800</td><td class="dd"></td><td class="clue">The type of SuiteAgents tool that sends emails, updates records, or triggers transactions — and requires human approval before executing.</td><td class="ans">What are action tools?</td></tr> | |
| <tr><td class="val">$1,200</td><td class="dd"></td><td class="clue">What SuiteAgents does after autonomously gathering data and synthesizing a recommendation — before taking any consequential action.</td><td class="ans">What is: requests human confirmation (human-in-the-loop gate)?</td></tr> | |
| <tr><td class="val">$1,600</td><td class="dd dd-star">★</td><td class="clue">The honest answer when a customer asks whether SuiteAgents is available to deploy today — accurate and commercially positioned.</td><td class="ans">What is: SuiteAgents is coming in NetSuite Next; AI Connector enables custom agentic flows today; HITL controls are part of the design?</td></tr> | |
| <tr><td class="val">$2,000</td><td class="dd"></td><td class="clue">SuiteAgents sits at Level 4 of the maturity ladder, but this is what separates it from a Level 3 'Acting' capability.</td><td class="ans">What is: SuiteAgents coordinates multi-step workflows across systems — not just a single tool call or transaction?</td></tr> | |
| </tbody> | |
| </table> | |
| <!-- ══════════════════════════════════════ | |
| FINAL JEOPARDY | |
| ══════════════════════════════════════ --> | |
| <h2>Final Jeopardy</h2> | |
| <div class="fj-box"> | |
| <div class="fj-label">Category</div> | |
| <p style="font-weight:bold;margin-bottom:1rem;">NetSuite AI Positioning</p> | |
| <div class="fj-label">Clue (read aloud after wagers are locked)</div> | |
| <p class="fj-clue">A prospect's CFO says: "I'm open to AI in our ERP, but I need to know that my team stays in control of financial decisions and that every AI action is auditable." Give the complete positioning response that addresses both concerns simultaneously — in one sentence.</p> | |
| <div class="fj-label">Correct Answer</div> | |
| <p class="fj-answer">What is: NetSuite uses deterministic controls around probabilistic reasoning — AI provides the contextual intelligence while rules-based guardrails ensure every action is auditable and high-risk decisions always require human approval?</p> | |
| <p style="margin-top:1rem;font-size:12px;color:#777;">Accept any answer that correctly names the principle and addresses both auditability and human control. Exact wording not required.</p> | |
| </div> | |
| </body> | |
| </html> |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment