← Back to curriculum

Module 8 — Agentic AI

Welcome to Module 8

What agentic AI means, interview focus areas, prerequisites from Module 7, and the travel planner project overview.

~25 min read + exercises

Welcome to Module 8 — agentic AI

Before we begin

Module 7 gave you RAG chatbots — one model, retrieved context, one answer. Agentic AI adds loops: the model decides which tools to call, observes results, and continues until a goal is met.

Interview focus: agents, tool calling, MCP, ReAct, evals, multi-agent orchestration, production design.

Figure

Module 8 at a glance

Module 8 — agentic AI1WelcomeModule 82Agentloop3ToolsAPIs4ReActplan5MCPctx6Multiagents7Evalstest8Designprod9Quizcheck10Projecttravel
Agents through system design, quiz, then a multi-agent travel planner with real APIs.

What Module 8 covers

TopicWhat you will understand
AgentsLoops, goals, vs plain LLMs and fixed workflows
ToolsFunction calling, schemas, execution
Planning vs executionReAct, specialized roles
MCP & contextTool servers, prompt assembly, memory
Multi-agentLangChain, LangGraph orchestration
EvalsLLM-as-judge, SME sets, regression gates
System designDeploy, scale, observability, guardrails

Key terms (read once)

Lessons use industry shorthand. If a word is new, check here first — each lesson also explains it again in context.

TermPlain English
DAGDirected Acyclic Graph — a fixed flowchart of steps (A → B → C) with no loops back; the engineer decides every branch in advance
WorkflowCode that runs steps in a predetermined order (often a DAG)
ReActReason + Act — the model alternates thinking in text and calling tools (Lesson 3)
MCPModel Context Protocol — a standard way to plug files, APIs, and databases into agents (Lesson 5)
HandoffOne agent passing a structured summary to another — not the entire chat history
OrchestrationCoordinating multiple LLM calls, tools, and shared state in the right order
EvalA repeatable test case + scoring rule (like a unit test for your AI feature)
SMESubject-matter expert — someone who knows the domain and writes gold reference answers
LLM-as-judgeA second model scores the first model’s output against a rubric
CIContinuous integration — automated tests that run when you push code (e.g. GitHub Actions)
JSONLJSON Lines — a text file with one JSON object per line (easy to append eval cases)
FixtureCanned fake API response used in tests so results are predictable
IdempotentSafe to retry: running the same action twice does not duplicate charges or records
Stateless (server)The app server stores no session in RAM; session data lives in Redis or a database
AllowlistOnly listed tools, paths, or actions are allowed; everything else is blocked
PIIPersonally identifiable information — email, passport number, home address, etc.
TTLTime to live — how long data is kept before it expires (cache, session, memory)
p95 latency95% of requests finish faster than this — a common “typical slow case” speed metric
Circuit breakerAuto-pause a feature when cost or errors cross a threshold (like tripping a fuse)
RegressionA change that breaks something that used to work — evals catch regressions before users do
SSEServer-sent events — the server pushes progress updates to the browser over HTTP
stdioStandard input/output — a local process the agent talks to via pipes (common for dev MCP servers)

Already from earlier modules: RAG (retrieve docs then answer), LLM (large language model), token (a chunk of text the model reads), embedding (a list of numbers representing meaning).


Before you start

Required: Module 7 RAG project or comfort with LLM APIs and prompts.

For the project:

  • LLM API with tool/function calling support
  • Weather API key (e.g. OpenWeatherMap) — free tier OK
  • Optional: maps/geocoding API
  • pip install langgraph langchain-openai (optional but recommended)

Ready?

Lesson 1 — What is an AI agent?