AI Essentials on AWS: Become Certified

Compare decision trees, linear and logistic regression, and neural networks to understand how model internals affect outputs and favor interpretable trees. Use accuracy to assess a plant disease image classifier.

Explore the fundamentals of gen AI and foundation models, including tokenizers, embeddings, and transformers, and learn how AWS infrastructure, zero-trust security, and the CIA triad enable secure AI applications.

Explore how foundation models generate text, images, code, video, and embeddings, and compare VAEs, GANs, and autoregressive models for data quality and task relevance.

Explore data retention and compliance requirements for GDPR, HIPAA, and SOC 2, automate monitoring with AWS config and Audit Manager, and implement envelope encryption with KMS and CMK.

Explore how GANs generate realistic data through a back-and-forth between a generator and discriminator, and how VAEs use an encoder, latent space, and decoder to create and interpolate new images.

Explore AWS generative AI tools and services like Bedrock and SageMaker to train, deploy, and run inference for text, image, and code generation.

Explore Amazon Bedrock's use cases: text generation, virtual assistants, text and image search, text summarization, and image generation, all powered by foundational models from multiple providers.

Explore Amazon Bedrock in a hands-on lab, using the text playground to compare models, adjust randomness and length, and learn model access, pricing, and serverless workflows.

Master hands-on image generation with AWS Bedrock, using an image generator model and prompts. Learn about access to foundation models, guardrails, and cost management in Bedrock.

Fine-tune foundation models with luxury watch data to personalize product descriptions, recommendations, and virtual assistance, boosting conversion, order completion, and repeat purchases.

Explore prompt engineering for foundation models and Amazon CodeWhisperer, craft specific, concise prompts with cross-file context and comments to improve code recommendations and unit tests.

Explore prompt engineering for guiding pre-trained language models with clear instructions and inputs to improve outputs without fine-tuning. Compare prompts to fine-tuning, highlighting lower resource needs and iterative improvement.

Learn to identify adversarial prompts and prompt misuses, apply input sanitization and guardrails, and prevent prompt injection and prompt leakage to ensure responsible AI behavior.

Explore Amazon Comprehend, a pre-trained NLP service that performs sentiment analysis, entity recognition, and text classification, with options for custom models and document support, enabling seamless document processing and insights.

Explore real-time and batch use of Amazon Comprehend to extract entities, key phrases, language, PII, sentiment, and targeted sentiment, then set up analysis jobs using S3 data and IAM roles.

Explore Amazon Textract, a fully managed machine learning service that extracts data from documents, handwriting, tables, and forms, preserving structure and enabling automated IDP pipelines with confidence scores.

Explore Amazon Rekognition, a cloud-based deep learning service for image and video analysis that offers pre-trained models for objects, faces, text, unsafe content, face liveness detection, and custom labels.

Explore Amazon Rekognition through a hands-on demo of label detection, image properties, moderation, and celebrity recognition, then analyze faces, age ranges, and text in images.

Set up Amazon Kendra by creating an index, adding data sources, and configuring IAM roles to enable enterprise search with connectors to S3 and Confluence.

Explore how bias and fairness emerge from historical data and how to mitigate them. Examine transparency, accountability, privacy, and environmental costs, and learn practical governance, audits, and privacy-preserving solutions.