Agentic AI Foundations / Glossary & Cheat Sheet
← Course Hub
TAKE-HOME REFERENCE

Glossary & Cheat Sheet

Every term from the workshop in plain English — agentic AI, Amazon Quick, and banking/finance shorthand. Searchable. Print it.

TermMeaning
AgentA system that perceives input, reasons over multiple steps, and takes action — closing the loop a plain chatbot leaves open
Perceive → Reason → ActThe core loop every agent runs, regardless of complexity
Tool UseAn agent's ability to call an external function (API, database, calculator) and use the result
Chain-of-Thought (CoT)Reasoning broken into visible intermediate steps rather than a single leap to an answer
MultimodalAbility to process inputs beyond text — scanned documents, images, screenshots
Workflow ArchitectureA fixed, deterministic sequence of steps with LLM reasoning inside each — predictable and auditable
Autonomous ArchitectureThe agent decides its own next step at each turn without a fixed script
Hybrid ArchitectureA deterministic skeleton with autonomous reasoning inside individual steps — the most common finance pattern
Multi-Agent ArchitectureSeveral specialized agents coordinated by an orchestrator
ChainingWorkflow pattern where each step depends on the prior step's output, run in fixed order
ParallelizationWorkflow pattern where independent sub-tasks run simultaneously, then results are merged
RoutingWorkflow pattern where a classifier step sends a request down the correct path
OrchestrationWorkflow pattern where a coordinator agent spawns and manages sub-agents toward a larger goal
SpaceA container holding documents and scoping an agent's knowledge — "the kitchen"
Chat AgentA saved, reusable prompt configuration with persona, instructions, knowledge, actions, and suggested prompts — "the recipe"
FlowA no-code, multi-step wrapper around one or more Chat Agents, schedulable and shareable — "the doorbell"
MCP ConnectorModel Context Protocol integration giving an agent live tool access to external systems — "the oven"
Quick ResearchAn AI-led, multi-step investigation feature that produces a cited report
Pilot CanvasCapstone deliverable — a structured one-pager for a proposed real agentic pilot, scored on autonomy/tools/guardrails/impact
Rule of ThreeHeuristic: typed the same prompt 3 times → save it as an agent
Set-and-Forget TrapFailure mode — an agent built once, never revisited as policy changes, quietly drifting out of compliance
Scope Creep SpiralFailure mode — one agent slowly absorbing too many responsibilities until its behavior is unpredictable
Ghost AgentFailure mode — an agent built once, shared, then silently abandoned
Audit-DefensibilityThe finance-specific design pattern: named output sections, traceable numbers, flag-only behavior, cited sources
Flag-only behaviorAn agent's guardrail that it can identify/classify an issue but never auto-resolve, approve, or clear it
Human-in-the-loopA required checkpoint where a person must review or approve before an action is final
Break (reconciliation)A line item where two records (e.g., GL and bank statement) don't match
Disposition memo/noteA drafted explanation of how an exception (break, NCR-style finding) should be classified and handled, for human sign-off
KYC / AMLKnow Your Customer / Anti-Money Laundering — the due-diligence and monitoring controls referenced throughout finance agent guardrails
Maker-CheckerA control requiring the person who initiates an action and the person who approves it to be different individuals