Production Deployment and Advanced Patterns

本课程是 Build Powerful AI Agents with OpenAI Tools 专业证书的一部分

位教师：Professionals from the Industry

访问权限由 Coursera Learning Team 提供

3个模块

深入了解一个主题并学习基础知识。

中级等级

推荐体验

1 周完成

在 10 小时一周

灵活的计划

自行安排学习进度

3个模块

深入了解一个主题并学习基础知识。

中级等级

推荐体验

1 周完成

在 10 小时一周

灵活的计划

自行安排学习进度

要了解的详细信息

可分享的证书

添加到您的领英档案

作业

6 项作业

授课语言：英语（English）

了解顶级公司的员工如何掌握热门技能

了解关于 Coursera for Business 的更多信息

Petrobras, TATA, Danone, Capgemini, P&G 和 L'Oreal 的徽标

积累 Software Development 领域的专业知识

本课程是 Build Powerful AI Agents with OpenAI Tools 专业证书专项课程的一部分

在注册此课程时，您还会同时注册此专业证书。

向行业专家学习新概念
获得对主题或工具的基础理解
通过实践项目培养工作相关技能
通过 Coursera 获得可共享的职业证书

该课程共有3个模块

The Production Deployment and Advanced Patterns course equips learners with the skills to transform AI agents into professional-grade applications ready for real-world use. Starting with user interface development, learners build interactive agent interfaces using Streamlit for rapid prototyping and Flask APIs for scalable integration across platforms. They then explore advanced UI patterns such as streaming responses, interactive visualizations, responsive design, and effective expectation management to create seamless agent experiences.

The course progresses to packaging and distributing agents through the GPT Store, emphasizing testing, optimization, deployment, and monitoring strategies for reliability and growth. Learners also master enterprise-grade integration, including role-based access control, audit logging, compliance documentation, and cost-performance management at scale. Finally, the course introduces DSPy, shifting from manual prompt engineering to programmatic, self-optimizing agent development with embedded ethical safeguards and continuous improvement pipelines. By the end, learners will have the skills to build secure, scalable, and future-proof AI systems.

As a Senior Consultant at Praxis AI, your mandate is to rescue "CityScout," a fragile travel agent prototype that crashes under real-world use. You will take a junior developer's messy, stateless Command Line Interface (CLI) script and refactor it into a robust production architecture, mastering the critical transition from Scripting to Application Engineering. Your mission is to implement State Persistence to stabilize the web interface, build a Headless Flask API to support mobile integration, and engineer Streaming Responses to solve latency issues, ensuring the system can handle concurrency and errors at scale.

涵盖的内容

3个视频5篇阅读材料3个作业3个非评分实验室

3个视频总计12分钟

1.1 Managing Session State in Streamlit 5分钟
1.5 Building a Flask REST API for Multi-User Sessions 3分钟
1.10 Building Real-Time LLM Token Streaming 4分钟

5篇阅读材料总计49分钟

1.2 The Stateless Web and Session Management 12分钟
1.6 RESTful Agent Design 10分钟
1.7 Specification-Driven Development: Building with Precision 5分钟
1.12 Spec-Driven Development: The Verification Layer 12分钟
1.15 Beyond Localhost: Distributing Agents via the GPT Store 10分钟

3个作业总计120分钟

1.14 Module 1 Assessment 60分钟
1.4 State Desync 30分钟
1.9 Endpoint Forensics 30分钟

3个非评分实验室总计120分钟

1.3 The Interface Refactor 30分钟
1.8 The API Builder 45分钟
1.13 The UX Polish 45分钟

Vitality Health has deployed TriageBot as a pilot program, but an audit reveals critical security gaps. Patient data (PHI) is being logged in plain text, there is no access control, and a billing spike shows the API is being abused. You must implement enterprise-grade security before the HIPAA auditors return. You will implement Role-Based Access Control (RBAC) to separate Doctors from Patients, sanitize all logs to remove Protected Health Information (PHI), and rate-limit the API to prevent DDoS attacks on critical infrastructure.