Building Generative AI Chatbot Applications
Contact us to book this course
Curriculum
Generative AI
Delivery methods
On-Site, Virtual
Duration
3 days
This course builds upon knowledge gained in the Introduction to Building Generative AI Applications course and prepares you to build conversational AI solutions such as chatbots and virtual assistants. The course consists of presentations, demos, and hands-on labs where students work with generative AI (GenAI) models and services, and build realistic GenAI chatbot applications.
Learning objectives
After successfully completing this course, you will be able to:
- Identify opportunities to leverage GenAI chatbots to address new business opportunities
- Understand key concepts specific to conversational GenAI solutions
- Evaluate various technology offerings and select tools best suited for their use cases
- Design solutions using best practices for security, safety, scalability, and cost
- Begin building GenAI chat applications from scratch, or using managed platforms
Who should attend
This course is designed for developers, solution architects, and others involved in the design, development, and operation of software systems.
Course outline
- Hands-On Tour of Real-World Chatbot Applications
- How Do Generative AI Chatbots Work?
- What Problems Do They Solve?
- How Do Chat Applications Differ From Basic GenAI Text Generation Applications?
- Data Sources
- Chat-Focused Models
- Fine-Tuned Models
- Embeddings and Vector Databases
- Application Architecture
- Chat History
- Client Sessions
- Integrations
- Technology Choices
- Technology Stacks for Custom-Built Applications
- Managed Chatbot Platforms
- Vertex AI Agent Builder and Dialogflow CX
- Microsoft Power Virtual Agents
- Design Best Practices
- Security
- Safety
- Scale
- Cost Management
- What Is LangChain, and Why Use It?
- LangChain Concepts
- Models
- Prompts
- Memory
- Indexes
- Chains
- LangChain and Multi-Turn Chat
- LangChain and Embeddings
- LangChain and Conversational Retrieval Chains
- Building Chatbots with a Standard LLM Backend
- Prompt Composition
- API Parameters
- Building Chatbots with a Fine-Tuned LLM Backend
- Best Practices
- Limitations
- Building Chatbots Using the Retrieval-Augmented Generation (RAG)
- Motivations
- Architecture
- Implementation
- Motivation
- Vertex AI Agent Builder
- Datastores
- Out-of-Box Dialogflow CD Functionality
- Expanding on the Core Application
- Microsoft Power Virtual Agents
- Datastores
- Integrating OpenAI Services
- Expanding on the Core Application
