Schița de curs

Introduction to Agentic AI

  • Defining agentic AI and its relationship to traditional AI systems
  • Overview of reasoning, memory, and goal-driven architectures
  • Key use cases and industry applications

Core Concepts and Design Patterns

  • The agent loop: perception, reasoning, and action
  • Single-agent vs. multi-agent systems
  • Environment interaction and tool invocation

Prompt Engineering Fundamentals

  • Designing effective prompts for reasoning and task decomposition
  • Using examples, constraints, and roles for better control
  • Debugging and iterating prompts systematically

Building Simple Agentic Workflows

  • Implementing an agent loop in Python
  • Integrating with APIs and simple tools
  • Managing agent state and memory

Responsible Design and Safety Practices

  • Ethical considerations and responsible use of agents
  • Bias, transparency, and accountability in AI systems
  • Access control, data protection, and content safety

Hands-on Project: Designing a Responsible Agent

  • Defining the problem scope and objectives
  • Developing the prompt and control logic
  • Testing, refining, and evaluating agent behavior

Summary and Next Steps

Cerințe

  • Înțelegere de bază a conceptelor AI sau de învățare automată
  • Familiarizarea cu sintaxa și scripting-ul Python
  • Experiență în lucru cu date sau aplicații bazate pe API

Audientă

  • Cercetători de date noi în dezvoltarea AI agentic
  • Ingineri ML juniori care exploră aplicările arhitecturilor agentiale
  • Manageri tehnici care doresc să înțeleagă principiile de design și siguranță ale agentelor
 14 ore

Numărul de participanți


Pret per participant

Mărturii (3)

Cursuri viitoare

Categorii înrudite