Category: Technical How-to

Fine-tuning large language models (LLMs) is crucial for leveraging their full potential across industries. Quantiphi unveils how fine-tuning supercharges LLMs to deliver domain-specific AI solutions that redefine possibilities. From personalized healthcare to precise financial predictions and streamlined legal reviews, fine-tuned models offer transformative value and unleash the power of customized, efficient, and responsible generative AI deployments.

A content delivery network (CDN) caches content closer to users and reduces load times. Monitoring CDN performance is crucial for optimizing user experience, and this post demonstrates building an Amazon QuickSight dashboard with generative AI capabilities to gain insights from Fastly CDN logs in Amazon S3. It covers configuring real-time log streaming, using AWS Glue to catalog data, accelerating dashboard creation using natural language, and creating rich data stories for stakeholders.

Integrating security scanning into infrastructure as code (IaC) allows detecting misconfigurations early in development. This post explores using the KICS plugin from Checkmarx with AWS CDK. Learn how KICS performs static code analysis to find vulnerabilities and compliance issues in IaC and scans the CDK code to report findings, failing the command for high-severity issues. This shifts security left, mitigating risks in deployed infrastructure.

As data becomes increasingly valuable, robust security measures are critical. This post reviews how Protegrity’s tokenization integration with Amazon Redshift Dynamic Data Masking enables organizations to effectively protect sensitive data. It provides an overview of key concepts like Protegrity Vaultless Tokenization and Redshift Dynamic Data Masking, including code examples to better safeguard sensitive information within their Redshift data warehouse both at rest and in use.

Amazon EC2 Mac instances provide Apple hardware to develop for iOS and macOS at scale. Dependencies for building apps require privileged access, normally needing manual user approval. By enrolling EC2 Macs into Mobile Device Management (MDM) with Jamf Pro, administrators can automate remote configuration and software deployment without per-instance interaction. After launching an instance and installing required profiles with Jamf, users should create an AMI to retain the instance state.

ServiceNow and AWS have collaborated to bridge language barriers in global workforces. Using AWS services like Amazon Translate and Amazon Comprehend, the AWS Translate Spoke for ServiceNow Flow Designer enables automatic translation of text into employees’ native languages. By demonstrating how the AWS Translate Spoke can translate knowledge articles, this post explains how ServiceNow customers can easily build multi-language workflows to serve global users.

Leveraging AWS Lambda response streaming functionality enables progressive data delivery from large datasets, enhancing web application performance. This post compares a traditional API implementation to a Lambda streaming API, demonstrating reduced time-to-first-byte latency and quicker, more dynamic client-side rendering. By incrementally sending data, response streaming eliminates lag from waiting on entire dataset transfers, vastly improving user experience for data-intensive applications.

As companies grow, data can become fragmented across business units, hindering insights. PwC built an enterprise data solution on AWS to centralize data and streamline ingestion from diverse sources into a consolidated data lake using native AWS services and configurations. This reduces bespoke data collection, freeing up time for analysis and innovation with consistent, up-to-date data sets. The expandable solution accelerates stand up and bridges silos to unlock operational insights across the business.

Generating high-quality synthetic data protects privacy and augments scarce real-world data for training machine learning models. This post shows how to integrate the Gretel synthetic data platform with Amazon SageMaker Pipelines for a full ML workflow. Gretel’s integration with SageMaker Pipelines in a hybrid or fully managed cloud environment enables responsible and robust adoption of AI while optimizing model accuracy. With Gretel, data scientists can overcome data scarcity without compromising individuals’ privacy.

Companies often struggle to understand customer sentiment in call center data, posing barriers to responding to emotions in real-time. SuccessKPI uses natural language understanding and machine learning on a large dataset to enable sentiment prediction. This helps avoid bias of human reviewers and attain objective results when analyzing customer feedback. SuccessKPI offers capabilities like sentiment by channel, time, quarter, and entity to transform customer experiences.