AWS AIP-C01: Amazon Bedrock Cost Optimization
Amazon Bedrock cost depends on more than which foundation model an application invokes. Input tokens, output tokens, prompt caching, inference routing, provisioned capacity,…
Read articleEDITORIAL TOPIC / EXAMTOPICS EDITORIAL
AI & Data.
Artificial intelligence, machine learning, data systems and analytics.
PUBLISHED COVERAGECurated by subject
AWS AIP-C01: Agentic AI Patterns with Amazon Bedrock
Agentic AI applications differ from ordinary prompt-response systems because the model can plan, call tools, retrieve context, preserve memory, and continue through several…
Read articleAWS DEA-C01: Kinesis Data Streams vs Firehose
Amazon Kinesis Data Streams and Amazon Data Firehose solve different streaming problems. Kinesis Data Streams is a durable real-time event stream for applications…
Read articleAWS AIP-C01: Building RAG Applications with Amazon Bedrock
Retrieval-Augmented Generation combines a foundation model with external evidence so responses can use enterprise documents, product knowledge, policies, support content, or other domain…
Read articleAWS DEA-C01: Lake Formation Permissions and Data Governance
AWS Lake Formation provides centralized governance for data lakes by adding fine-grained permissions over data stored in Amazon S3 and metadata in the…
Read articleAWS AIP-C01: CI/CD for Generative AI Applications
Generative AI applications need CI/CD, but the release artifact is broader than ordinary application code. A production change can involve prompt templates, model…
Read articleAWS DEA-C01: Data Quality and Lineage in AWS Pipelines
Data pipelines can complete successfully while producing incorrect, stale, duplicated, or semantically broken output. That is why production data engineering needs two complementary…
Read articleAWS DEA-C01: EMR vs Glue for Large-Scale Data Processing
Amazon EMR and AWS Glue can both run distributed data processing on AWS, but they optimize for different levels of control. EMR gives…
Read articleAWS DEA-C01: Building Data Lakes on Amazon S3
Amazon S3 is a common foundation for AWS data lakes because it provides durable object storage, virtually unlimited scale, flexible prefixes, lifecycle controls,…
Read articleAWS DEA-C01: Glue Jobs, Crawlers & Data Catalog
AWS Glue combines metadata management and serverless data integration so teams can discover datasets, define schemas, transform data, and make results available to…
Read articleNVIDIA NCA-AIIO: Storage Design for AI Training Workloads
AI training pipelines can consume enormous amounts of data, but raw capacity is only one storage-system requirement. Training also needs enough read throughput,…
Read articleNVIDIA NCA-AIIO: Sizing GPU Systems for Training and Inference
GPU sizing is not a simple question of model parameter count divided by device memory. Training and inference consume memory and compute in…
Read articleNVIDIA NCA-AIIO: Securing Enterprise AI Infrastructure
Enterprise AI infrastructure combines high-value data, privileged compute, proprietary models, large-scale storage, specialized networking, container orchestration, secrets, management planes, and expensive accelerators. That…
Read articleNVIDIA NCA-AIIO: Networking for Multi-GPU AI Clusters
Multi-GPU AI performance depends heavily on communication. Large models exchange gradients, activations, parameters, and optimizer state across devices, while inference systems move request…
Read articleNVIDIA NCA-AIIO: GPU Scheduling and Resource Isolation
Shared GPU infrastructure creates a scheduling problem: many teams want expensive accelerators, but their workloads differ in duration, memory demand, priority, latency sensitivity,…
Read article