This repository provides a curriculum for learning Agentic AI, focusing on the Dapr Agentic Cloud Ascent (DACA) design pattern. It targets AI developers and AgentOps professionals aiming to build scalable, resilient multi-agent systems, particularly addressing the challenge of handling millions of concurrent agents. The core benefit is a structured learning path from foundational concepts to planet-scale deployment.

How It Works

DACA integrates the OpenAI Agents SDK for agent logic, the Model Context Protocol (MCP) for tool use, and the Agent2Agent (A2A) protocol for inter-agent communication. It leverages Dapr for distributed capabilities, enabling stateful actors and event-driven workflows. The pattern promotes cloud-native deployment on platforms like Kubernetes or serverless containers, emphasizing modularity and context-awareness for efficient scaling.

Quick Start & Requirements

• Installation: Primarily through course materials and local development setups using tools like Rancher Desktop.
• Prerequisites: Successful completion of AI-101 (Modern AI Python Programming). Subsequent courses (AI-201, AI-202, AI-301) build upon each other, requiring specific knowledge of Python, FastAPI, containerization (Docker, Kubernetes), Dapr, and various databases/messaging systems.
• Resources: Local development can utilize Rancher Desktop. Cloud-based deployments may involve cloud credits or free tiers.
• Links:
  • AI-201 Video Playlist: https://www.youtube.com/playlist?list=PLh_1_1_2_3_4_5_6_7_8_9_0 (Note: This is a placeholder URL as the actual playlist is not provided in the README.)

Highlighted Details

• Focuses on the DACA design pattern for building planet-scale multi-agent systems.
• Addresses the challenge of managing 10 million concurrent AI agents, with detailed arguments for feasibility using Kubernetes and Dapr.
• Compares various AI agent frameworks, advocating for OpenAI Agents SDK's balance of simplicity and control for most use cases.
• Structured curriculum covering AI-201, AI-202, and AI-301 courses, with detailed weekly breakdowns.

Maintenance & Community

• Developed by Panaversity as part of their Certified Agentic & Robotic AI Engineer program.
• The project's primary goal is educational, with a focus on training developers in Pakistan and globally.

Licensing & Compatibility

• The README does not explicitly state a license.

Limitations & Caveats

• Direct benchmarks for 10 million concurrent users with Dapr/Kubernetes in an agentic AI context are not provided.
• Achieving planet-scale deployment requires significant infrastructure investment, which may be prohibitive for low-budget scenarios, though the curriculum suggests simulation and cloud credits as alternatives.