Code-first Jupyter tutorials for building, deploying, and scaling production AI agents.
Move from prototype to enterprise GenAI deployment using hands-on code examples.
This repository is an open-source educational collection of code-first Jupyter Notebooks for building AI agents. It serves developers, engineers, and architects looking to transition GenAI prototypes into production systems. You can run individual notebooks to implement specific architectural components like vector memory or Docker deployments.
The project delivers 28 end-to-end tutorials covering orchestration frameworks like LangGraph, Model Context Protocol runtimes, and real-time web search APIs. It includes modular code for security guardrails, GPU cloud scaling, browser automation, and observability tools. Each module lives in its own directory with runnable code files and setup scripts.
It suits software engineers, AI architects, and developers who need practical code patterns for deploying LLM workflows. It does not suit non-technical users who want a visual, no-code agent builder or a pre-packaged turnkey application.
Every tutorial contains standalone code designed to execute in local or cloud Jupyter environments.
Implement multi-agent coordination and durable execution patterns using frameworks like LangGraph.
Package agents with FastAPI endpoints, Docker containers, and GPU cloud infrastructure.
Apply tool-calling authorization controls via OAuth2 and human-in-the-loop validation steps.
The tutorial code itself is free to access and modify, though no repository license is explicitly attached. You must cover your own operational costs, including LLM API usage, cloud hosting fees, and GPU compute instances required by specific notebooks.
896 open remote AI-training roles, updated November 16, 2026; most pay $20–$175 an hour. Zekai lists them free; the application goes to the employer’s platform.
No, it is a curated set of educational tutorials and reference code designed to help you construct your own agent stacks.
Yes, most tutorials require third-party LLM, vector database, or web search API keys to execute.
See Zekai first in Google