Category: Software

Amazon Athena allows analyzing petabytes of data directly with SQL queries or via analytics tools like ThoughtSpot. This post covers creating an IAM access key, two Amazon S3 buckets (one for data storage, one for query results), creating an Amazon Athena table from sample data using an AWS Glue crawler, and establishing a connection between Athena and ThoughtSpot. With the connection, ThoughtSpot can search and visualize the Athena data using AI capabilities.

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.

SaaS relies on the cloud for efficient delivery, but faces data challenges like tenant isolation, high storage costs, and compliance requirements. Amazon FSx for NetApp ONTAP solves these by providing multi-tenancy, storage efficiencies, persistent Kubernetes storage, and more. It enables SaaS providers to enhance agility while managing data availability, protection, security and costs. Arcesium reduced dev/test refreshes by 80% and overall costs by 46% using FSx for ONTAP.

Pinecone has developed a novel serverless vector database architecture optimized for AI workloads like retrieval-augmented generation. Built on AWS, it decouples storage and compute and enables efficient intermittent querying of large datasets. This provides elasticity, fresher data, and major cost savings over traditional architectures. Pinecone serverless removes bottlenecks to building more knowledgeable AI applications economically at scale on AWS.

The Synopsys.ai electronic design automation (EDA) suite on AWS harnesses AI to optimize chip design. A key component is Synopsys DSO.ai, using reinforcement learning to enhance power, performance, and area. Benefits include faster optimization, better engineer productivity, design reuse, and faster process node migration. Deploying on AWS ParallelCluster provides auto-scaling, elasticity, and fast setup for massive EDA workloads. Testing showed 20% lower power, timing closure improvements, and significant cost savings.

Amagi provides cloud broadcast solutions with tiered redundancy offerings to mitigate risks for broadcasters during disruptive events. Its CLOUDPORT platform supports hot, warm, and cold disaster recovery configurations spanning single/multi AWS Availability Zones and regions for varying recovery times. Exhaustive tuning, testing, and regular DR drills validate the failover capabilities aligned with broadcaster requirements for cost-effective business continuity.

As organizations move to the cloud, securing workloads, applications, and data becomes critical. This post discusses using Zscaler Zero Trust Exchange to provide comprehensive cloud security. Key security capabilities highlighted include SSL inspection, data loss prevention, and URL filtering policies tailored for cloud workloads. The post provides guidance on design considerations and implementing baseline security policies within the Zscaler platform.

Many organizations want to migrate from legacy BI tools to Amazon QuickSight, but manual migration cab be costly. Wavicle’s QuickSight Migration Solution (QMS) automates and optimizes this process, analyzing existing dashboards, identifying duplication, automating conversion, and facilitating a seamless rollout. This reduces migration effort, costs, and risks so organizations can rapidly realize QuickSight’s self-service analytics, flexibility, and cost savings.

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.

The Hugging Face Platform provides no-code and low-code solutions for deploying generative AI models on managed AWS infrastructure. Key features include Inference Endpoints for easy model deployment, Spaces for hosting machine learning apps, and AutoTrain for training state-of-the-art models without coding. Hugging Face is an AWS Generative AI Competency Partner whose mission is to democratize machine learning through open source, open science, and Hugging Face products and services.