Agentic AI using Programming

Duration8 weeks

FormatLive online + labs

LevelIntermediate–Advanced

FeeEnquire

About the course

Agentic AI using Programming is for engineers who need systems that do more than answer a question: they plan, call tools, inspect results, and retry. You will implement agents with LangChain, LangGraph, and related patterns, then wrap them with logging, human-in-the-loop, and deployment basics so they are safe enough for a pilot.

Learning outcomes

Design agent loops with tools, memory, and explicit stop conditions
Orchestrate multi-agent or graph-based workflows
Add tracing, evaluation, and human approval gates
Deploy a small agent service with Docker and an API
Present a capstone that a hiring manager can run and review

Learning plan

Tool calling, ReAct-style loops, when agents overcomplicate a problem, and designing a narrow first agent.

Hands-on labs and projects

Research agent with search + document tools
Workflow agent that updates a ticket or spreadsheet via API
Capstone: domain agent with evals and a README a recruiter can follow

Technologies

PythonLangChainLangGraphCrewAIFastAPIDocker

FAQs

No. This track is programming-first. If you want workplace AI without writing production code, see Microsoft Copilot.

Related learning path

AI Engineer

From Python fluency to production agents

View path

Related courses

Enquire

Intermediate · 6 weeks

Retrieval-Augmented Generation (RAG)

Design enterprise knowledge assistants that retrieve from your documents, cite sources, and stay within policy.

View details →

Enquire

Beginner · 4 weeks

Generative AI Foundations

Build a practical foundation in large language models, prompting, evaluation, and responsible use — ready for workplace projects.

View details →

Back to all courses