Amazon Bedrock with Amazon Q Developer-Zero to Hero | Python

Discover Amazon Bedrock, a fully managed, one-stop platform unifying foundation models from multiple providers, with custom fine-tuning, seamless AWS integration, and use cases in content generation, chatbots, and data insights.

Explore Amazon Bedrock with Party Rock to build apps without credentials, try free, and compare models like Titan, Llama, and Stability AI to generate country-based menus and widgets.

Explore the Amazon Bedrock UI, including Playground and Bedrock Studio, and learn about foundation models, providers, prompts, building custom prompts, agents, safeguards, and text and image generation.

Enable model access in Amazon Bedrock through the AWS console, choose all models or specific ones, and provide organization and contact details for setup.

Explore post model access in AWS Bedrock after authorization. Some models, like Anthropic, Llama, and Meta, require AWS team approval, while normal use uses available models.

Learn to list all Bedrock foundation models using a Python boto3 client in a lambda function, guided by Amazon Q’s CodeWhisperer assistance.

Explore the chat playground to test multiple models, compare providers, and evaluate prompts, latency, and model metrics for informed selection.

Master the text playground in Amazon Bedrock with single-input, single-output interactions, contrasting it with chat workflows and showing how to run models and view available options.

Explore the image playground in Amazon Bedrock with Amazon Q, comparing Titan AI and Stability AI for text-to-image generation, previewing tiger and zoo scenes with configurable image counts.

Explore how Amazon Q developer acts as an AI-powered copilot for AWS applications, providing inline code suggestions, debugging help, security checks, and code generation across CLI, console, and IDE.

Discover how Amazon Q developer serves as a co-pilot for building AWS projects, diagnosing console errors, transforming code across Java versions, and integrating with IDEs and chats.

Demonstrates using the Amazon Q developer in the AWS console to access service details, billing data, and EC2 resources across regions. Explore cross-region settings and SageMaker basics.

Demonstrate invoking a bedrock model with Python in Lambda using Boto3, comparing code whisperer and Amazon Q developer workflows to generate a calculator function and handle errors.

Learn to use Amazon queue with Bedrock to generate text playground responses. Implement a three-step flow: declare the Bedrock runtime client with Boto3, invoke the model, print the output.

Configure a Lambda function using Boto3 to invoke Amazon Bedrock Titan for text prompts, grant Bedrock access, and test end-to-end text generation via the Bedrock runtime.

Explore bedrock text models with Claude Sonnet hands-on, comparing hard-coded prompts vs. test-case prompts, using lambda with boto3 to work with anthropic models like 3.5 sonnet and haiku.

Learn to create a python function and call bedrock's llama text model, using prompts, request and response examples, and permission settings to deploy and test in your environment.

Learn to troubleshoot Amazon Bedrock queue issues in Python, resolving no recommendation is available and incomplete or irrelevant results by using console queue, validating prompts, and ensuring correct region.

Learn to generate images with Amazon Titan using Bedrock's Titan image generator, implement text-to-image workflows in Lambda, and store outputs in S3 with boto3.

Learn to generate images with Stability AI's Bedrock image model in a Python lambda workflow, configure permissions for Bedrock and S3, encode outputs as base64, and store pngs in S3.

Understand what artificial intelligence is, its key features, and how machine learning, NLP, and computer vision power applications in healthcare, finance, and everyday life, including trends and deep learning.

Understand generative AI and how foundation models train on data to generate content. Explore applications in creative arts, healthcare, and gaming, and future trends toward realism and personalized education.

Explore opportunities and challenges of generative AI across healthcare, education, creative arts, and business automation. Learn ethical strategies—transparency, bias mitigation, fair data, audits, and ongoing R&D to guide responsible deployment.

Understand how artificial intelligence, machine learning, deep learning, and generative ai connect, with notes on neural networks and supervised, unsupervised, and reinforcement learning.

Differentiate foundational models from large language models by understanding that foundational models handle diverse data and modalities, while LLMs focus on text with billions of parameters and context-aware generation.

Master prompt engineering by crafting clear prompts with context settings and personas, and iteratively refining outputs from foundation models for tasks like customer support, content creation, and data analysis.

Explore the different types of prompt engineering for foundation models, including instruction-based prompts, few-shot learning prompts, chain-of-thought prompts, and persona-based prompts.

Explore how inference parameters control the randomness, diversity, and predictability of text produced by foundational models, and learn how temperature, top p (nuclear sampling), and top k sampling shape outputs.

Explore how temperature in generative AI controls creativity versus predictability, ranging from 0 to 1, with concrete storytelling examples and discussion of how outputs vary.

Learn to configure temperature in Amazon Titan Bedrock using a Python lambda function, adjusting 0–1 to control randomness and test prompts.

Explore how temperature shapes prompts when using Mistral on Amazon Bedrock with a Python lambda workflow. Configure a lambda function, adjust top_p and top_k, deploy, and test prompts.

Discover how the top parameter governs probability to balance predictable and creative outputs, and how temperature and model selection drive end-to-end text generation.

Explore how to configure top-p in a llama bedrock setup using Python and Boto3, test prompts, and deploy a function to generate controlled responses.

Explore top-p controlled generation with q developer on the anthropic cloud using bedrock models; learn setup, code with python and boto3, handle model access, prompts, and configuration.

Explore how top k configurations limit responses and balance randomness and predictability using temperature and top p, with hands-on examples and model availability considerations.

Hands-on with top k in Amazon Q and Bedrock teaches three steps: declare the runtime, create the code, and call it, while illustrating model support and value ranges 0–500.

Hands-on session demonstrates building a top k query on Amazon Bedrock using Mistral AI with Lambda and Boto3, covering top k from 1 to 200 with default 50.

Learn to compute the most probable answer by tuning temperature, top p, and top k across Bedrock, Mistral AI, and cloud models, with hands-on AWS Lambda integration.

Identify the access denied error when bedrock lacks a permission set. Update the bedrock permission set to grant bedrock full access and test invoke operations with a python lambda function.

Resolve lambda timeout errors in Amazon Bedrock workflows by increasing the timeout, configuring bedrock permissions, and invoking the Amazon Titan Bedrock model with boto3, including deployment and testing steps.

Learn to troubleshoot validation errors in Amazon Bedrock by ensuring correct input formatting and request parameters, including proper prompt syntax for cloud models.

Resolve invoke modal error by requesting access or switching to an available model in bedrock, and verify permissions and configurations to test with supported models.

Explore additional configuration parameters for inference, including system prompts, max length, and stop sequences. Tune these options to control output beyond temperature, top P, and top k.

Explore how system prompts define a persona to guide model behavior across use cases. They are optional and can be applied to models, with examples like acting as a kid.

Learn to create and test a system prompt for Amazon Bedrock with Amazon Q Dev using Python and boto3, configure timeouts and permissions, and set model tone.

Select a model and set the maximum length to control how long the output should be; adjust to 200 words to see the effect on tokens and cost.

Configure maximum length and maximum token count in Amazon Bedrock to control input and output size across models like Titan and Express, optimizing output quality and cost.

Configure and test stop sequence in generative AI to halt output when a pattern or keyword is matched, using AI21 Labs models for real-time demonstrations.

Learn how to implement a stop sequence in Amazon Bedrock using Python to halt output when a regex is matched, with practical steps for configuration, testing, and deployment.

Implement safeguards in generative AI by applying guardrails to block harmful topics and watermark detection to flag AI-generated images, enabling HIPAA-compliant content filtering, usage monitoring, and access control.

Set up guardrails to restrict topics and enforce policy across an organization. Configure block messages, topics, PII controls, and test with different models to validate responses.

Learn to use guardrails in the chat option, upgrade versions as policies evolve, and manage the working draft and versions for policy-aligned guardrails.

Learn watermark detection to verify if an image is ai generated, using Amazon Titan image generator and stability ai, noting detection works with Titan but not stability ai.

Explore builder tools like prompt management, agents, knowledge base, and prompt flow to configure foundation models for specific use cases.

Master prompt management with a builder tool, defining prompts, variables, and encryption, then compare models and tune prompts for artificial intelligence in manufacturing.

Learn to build a knowledge base in Amazon Bedrock by uploading organization documents to S3, selecting Titan embeddings, and configuring data sources and IAM roles; avoid root access.

Learn how IAM users provide long-term access with unique names and policies, while IAM roles grant temporary access for integrating AWS services like EC2 to access S3.

Explore how IAM policies in AWS define permissions for users, groups, and roles using JSON statements of actions, resources, and conditions; create and attach built-in or inline policies.

Learn to create an IAM user, set up a Bedrock knowledge base, sync data, and query it using an embedding model to answer policy questions.

Explore Amazon Bedrock agents as task helpers that perform data retrieval, content generation, and customer interactions. Connect them with APIs, databases, and AWS services to enable context-aware, scalable workflows.

Create and configure an agent to assist with tasks like hotel bookings and flight reservations using builder tools, knowledge bases, and guardrails; choose models (Anthropic, Amazon, Meta) and define actions.

Create and configure agents with the assistant by supplying prompts, building action groups for booking flights, hotels, and rental cars, and preparing and testing the agent.

Create an HR-based agent manually that answers employee queries using a knowledge base. Save the agent progressively, configure memory and encryption, enable code interpreter, and test HR policies.

Learn to add action groups and guardrails to a manual agent, configure user input for leaves with date and days parameters, and manage safety, memory, and model options.

Explore prompt flow as a drag-and-drop pipeline that shapes conversation paths, conditionally routing user inputs to fetch data and produce outputs for generative AI models.

Learn to design end-to-end workflows with prompt flow using drag-and-drop, integrate knowledge bases, and configure KMS encryption, permissions, and AWS Lambda functions, Amazon Queue, or Amazon Lex for scalable responses.

Build and manage prompt flows in PromptFlow by creating and configuring prompts, wiring topic and industry inputs, and generating outputs with the sonnet model for manufacturing AI applications.

Build end-to-end prompt flows from scratch, define prompts for a music playlist and holiday policy checks, and configure models, temperatures, inputs, and knowledge bases.

Build a condition flow in PromptFlow using the condition builder to test if the input cloud is AWS, and respond with AWS supported or other clouds not supported.