Agentic AI and MCP Servers from Scratch

What you'll learn

Build multi-step AI agents using Python, LangChain, and LangGraph

Connect agents to external tools and APIs using Model Context Protocol (MCP)

Implement Retrieval Augmented Generation (RAG) with vector databases

Create production-ready AI pipelines with error handling and fallbacks

Deploy agentic systems to cloud infrastructure

Monitor and evaluate agent performance with real-world metrics

Build custom MCP servers for any data source or REST API

Design agentic workflows for measurable business automation

Capstone Project

Deploy a Production Agentic Business Assistant

Build and deploy a multi-tool AI agent that answers questions about your data, books calendar events, generates reports, and escalates edge cases. Hosted on cloud infrastructure with monitoring and evaluation.

LangGraph multi-step agent

Custom MCP server connecting to 3 data sources

Deployment with Docker + CI/CD

Agent evaluation dashboard

Course curriculum3 modules · 9 lessons

Python and LLM Foundations

3 lessons

↓

Set up your AI engineering environment, learn Python for API calls and async processing, and master prompt engineering for reliable outputs.

Python Environment Setup and Your First API Call

Preview35 minutes

Working Directly with the Anthropic and OpenAI APIs

50 minutes

Prompt Engineering for Reliable, Repeatable Outputs

55 minutes

Agents, RAG and Tool Use

3 lessons

↓

Build LangChain pipelines, design multi-step LangGraph agents, and implement Retrieval Augmented Generation with vector databases.

LangChain Fundamentals: Chains, Memory and Routers

60 minutes

LangGraph: Multi-Step Agent State Machines

70 minutes

RAG: Retrieval Augmented Generation with Pinecone

65 minutes

MCP Servers and Production Deployment

3 lessons

↓

Build a custom MCP server, connect Claude to your business data sources, and deploy agents to cloud with monitoring and evaluation.

Building Your First MCP Server

75 minutes

Connecting Claude to External APIs via MCP

60 minutes

Deploying and Monitoring Agents in Production

80 minutes

About this course

Agentic AI is the next frontier: AI systems that don't just answer questions but take multi-step actions, use tools, query databases, browse the web, write code, and execute workflows. This course teaches you to build these systems from first principles. You'll start with Python fundamentals for AI engineering, then move through LangChain and LangGraph to build agents that reason, plan, and execute. You'll build a working MCP server that connects Claude to your business data — calendar, CRM, files, APIs — and deploy it to a cloud environment. The capstone is a fully functional agentic business assistant: it answers questions about your data, books appointments, generates reports, and escalates edge cases to a human. Real infrastructure, real deployment, real business value.

Requirements

Basic Python knowledge helpful but not required — fundamentals are covered
Interest in AI systems and workflow automation
Laptop with internet connection

Your instructor

Biruk A.

Senior AI Engineer · Adjibar

Built production LangGraph agents for federal contractors and commercial clients. Claude API certified developer. Contributor to open-source AI tooling.

4.9 instructor rating1,200+ students

Frequently asked questions

How much Python experience do I need?

Basic familiarity with Python is helpful but the first module teaches everything from scratch. If you can write a for loop, you're ready.

Do we use real cloud infrastructure?

Yes. The deployment module uses a real cloud VM (AWS or DigitalOcean). The course provides $50 in cloud credits to cover lab costs.

What makes this different from other AI courses?

This course is built around production patterns — not Jupyter notebooks. Every module produces something deployable. The capstone is a live agent URL, not a slide deck.

Is MCP the same as function calling?

MCP extends function calling into a standardized protocol for connecting AI to any data source. We cover both concepts and explain when to use each.

This course includes

30-day money-back guarantee

Verified certificate of completion

Community access for 6 months

Downloadable templates and resources

Lifetime access to all content

Flexible self-paced schedule

Looking for team training?

We offer private cohorts for teams of 5 or more. Custom curriculum, dedicated instructor, and group pricing.

Request Team Pricing

Related bootcamp track

Go deeper with a full instructor-led bootcamp. Includes live sessions, capstone coaching, and job placement support.

View bootcamp programs

Ready to start?

Invest in your skills today.

One-time payment. Lifetime access. 30-day money-back guarantee. Join 1,200+ professionals who have already enrolled.

Enroll Now — $299.00 Browse all courses