Скачать или смотреть Agentic AI Workshop 3 of 3: Evaluations, Security, and Production Deployment

In this final workshop of the 3-part series on Agentic AI systems, we cover the critical aspects of making agentic AI systems production-ready. You'll learn how to evaluate system performance, implement security measures, ensure user safety, and deploy your applications with confidence.

You'll learn how to:
Evaluate agentic AI systems using technical metrics and business KPIs
Measure task performance including faithfulness, relevance, and semantic similarity
Apply automated testing methods: LLM as judge, rule-based testing, and golden datasets
Secure applications against prompt injection and system exploitation
Implement guardrails to protect both users and systems
Assess app safety risks: content, behavioral, and contextual
Deploy portfolio projects using Gradio, Streamlit, and Hugging Face Spaces
Build production-grade systems with proper error handling and monitoring

Timestamps:
0:00 - Introduction and workshop series recap
2:30 - Introduction to evaluations and testing
6:07 - Task performance and technical metrics
13:11 - Security and robustness of agentic systems
16:47 - App safety and clean hands failures
24:16 - Deployment tools for portfolio projects
27:42 - Production deployment considerations
31:37 - Q&A on LangGraph and agent frameworks

Watch this workshop if you want to move beyond building basic agentic AI systems to creating secure, reliable, production-ready applications that you can confidently deploy in real-world environments.

This is Workshop 3 of 3 in the Agentic AI series. Workshop 1 covered foundational LLM workflows and RAG systems, and Workshop 2 focused on multi-agent architectures and design patterns with n8n.

This workshop is adapted from the Agentic AI Developer Certification Program by Ready Tensor - a 12-week, project-based program where you go far beyond this introductory session to design, build, and deploy real-world agentic AI systems.

In the full program, you'll progress from foundational LLM workflows and RAG systems to multi-agent architectures, MCP-based tool integration, evaluations, guardrails, and production-ready deployments.

Learn more and enroll here:
https://www.readytensor.ai/agentic-ai...

These Agentic AI workshops were delivered in partnership with the University of San Diego's Masters programs in Applied AI and Data Science. We've made them public so anyone interested in building agentic systems can learn the fundamentals.

About Ready Tensor:
Ready Tensor helps AI and ML professionals learn by building. Our platform combines certifications, hands-on projects, competitions, and community-driven learning to prepare you for real-world AI development.

Learn more: https://www.readytensor.ai/

If you found this workshop useful, like the video, subscribe to the channel, and let us know what agentic AI topics you want us to cover next.