Mastering GenAI on AWS

Explore the core concepts of generative ai, powering technologies, and how large language models transform industries, with Amazon Bedrock providing the tools to build gen AI applications for this course.

Unpack what gen AI is, how it works, and its impact today. Explore GANs, variational autoencoders, and transformer models like GPT and DALL·E, and examine training and ethics.

Explore the technologies behind gen AI, including machine learning foundations, deep learning models, and key algorithms such as supervised, unsupervised, reinforcement learning, CNNs, and GANs, with real-world applications.

Explore large language models and small language models, their data-driven training lifecycle from data to training to inference, and real-world tasks like translation, summarization, and code generation on AWS.

Explore Amazon Bedrock, a fully managed serverless service offering foundation models via a single API, featuring retrieval augmented generation and knowledge bases with private data sources.

Discover how to leverage AWS AI services, prebuilt models, and AI tools to build powerful applications, starting with setting up your AWS account and exploring the AI services landscape.

Explore the AWS AI ecosystem, from gen AI services like Amazon Q and Bedrock to tools such as Lex, Transcribe, Polly, Textract, and Comprehend.

Sign up for an AWS free account, enable MFA on root account, and configure billing alerts to control costs while exploring gen AI services like Amazon Q and Amazon Bedrock.

Set up the development environment for the course, covering Docker with VSCode, Jupyter notebooks, and the Lang Chain framework to enable coding and experimentation.

Set up a complete dev environment in VSCode with docker and dev containers, and use LangChain with Amazon Bedrock to build LMS apps with external data sources and notebooks.

Set up your genai development environment with VS code and Docker Desktop. Install the dev containers extension and configure dev containers with devcontainer.json and Docker file.

Master the LangChain framework for building LLM-powered apps, exploring chains, agents, memory, tools, and prompts. See modular design, provider support, and use cases like conversational AI and automated data processing.

Set up jupyter notebooks by running a docker container from the data science notebook image and creating intro.ipynb with the python 3.11.6 kernel. Prepare for lamp chain on Amazon Bedrock.

Gain hands-on experience with Amazon Bedrock, foundation models, and advanced prompt workflows, including prompt templates, LM chains, bedrock agents, knowledge bases, and bedrock studio.

Amazon Bedrock, a fully managed AWS service that provides access to foundational models, prompts, knowledge bases, prompt flows, Bedrock Studio, and guardrails with throughput control.

Launch your first hands-on with Amazon Bedrock and Cloud Instant, set up VS Code, obtain model access, and interact via a Jupyter notebook to explore text-based prompts and LM capabilities.

Explore building and using string prompt templates with Bedrock LM and LangChain prompts to format dynamic inputs, extend prompts with concatenation, and specify language for responses.

Hands-on guide to building chat prompt templates that fuse system and human prompts for multi-message LLM interactions. Use dynamic variables like domain, task, and question to tailor Python programming answers.

Explore LM chains as sequences of calls, including simple sequential, transform, and outer chains, and build a Bedrock LM chain using prompts and input data.

Explore zero-shot, one-shot, and few-shot prompts and how templates and examples shape llm responses in a customer service context, using an llm chain to generate professional replies.

Explore Amazon Bedrock agents to augment LLM workflows with live data from third-party tools, using a weather demo with Lambda and openweathermap.

Explore how LangChain agents on AWS Bedrock use zero-shot react description prompting to fetch a paper from archive.org and generate a structured summary through chain-of-thought reasoning.

Explore prompt management in Amazon Bedrock: use a library of prompts, create and test prompts with input variables, and deploy production versions for apps.

Explore Amazon Bedrock prompt flows to build workflows by chaining prompts from Prompt Management with foundational models and AWS services, creating end-to-end outputs like email drafts.

Build and configure Amazon Bedrock knowledge bases to power RAC workflows. Choose data sources like S3, set up embeddings and a vector store, and use prompts and agents.

Explore Bedrock Studio, a ssl-enabled web interface to build llm apps in aws bedrock; configure identity center access, permission boundaries, and roles, then create workspaces, prompts, and knowledge bases.

Discover Amazon Cube to build and customize ai applications with ease, set up your first app, manage data sources and access controls, and deploy real-world ai solutions.

Introduce Amazon Queue, a gen ai powered assistant embedded in the AWS console for querying services, code tasks, and private enterprise data with developer and business variants.

Create your first Amazon Q application, configure service roles and encryption, choose a retriever and index provisioning, and connect Identity Center for users and the web experience.

Configure data sources for the gen ai app by creating an s3 bucket, uploading files, and defining sync, filtering, and access options to index and ready data for use.

Customize the Amazon Q web experience with a personalized AI financial assistant, and learn to summarize contracts and query enterprise data stored in S3 using prompt engineering.

Configure admin controls and guardrails in the amazon cube app, including global and topic-specific controls, response settings with blocked words, and data source management.

Celebrate completing Mastering Gen AI on Amazon web services by applying in-depth Gen AI concepts with Amazon Badrock and Chain Framework, embracing experimentation and continuous learning to build real impact.