Matthew James Eddy的动态

?? Develop a Generative AI application using AWS. Check out this video for a brief demonstration of AWS Bedrock, showcasing how to create the RAG application. ?? Watch this video

???????? ?? - ?????????? ???????????????? ?????????????????? ???????????????? ???????? ?????????? ????!?? Excited to share my recent project: a Question Answering Solution powered by Microsoft Azure Cognitive Services! This hands-on experience explored how Azure’s AI capabilities simplify complex challenges and deliver real-world impact. ?? Project Highlights: ? Python Integration Seamlessly integrated Azure’s Question Answering Client for dynamic Q&A functionality. ? Custom Knowledge Base Built a tailored knowledge base to handle FAQs with efficiency and accuracy. ? Transparent Responses Displayed answers with confidence scores and source references for enhanced user trust. This project highlights the immense potential of Azure AI in building intelligent, scalable, and user-friendly Q&A systems. ?? Check out the code here: https://lnkd.in/gMjJM5dy Have you explored the power of Azure AI? I’d love to hear your thoughts and ideas. Let’s connect and innovate together! ?? #MicrosoftAzure #CognitiveServices #PythonProgramming #MachineLearning #CloudAI #AIInnovation #TechJourney 4o

I tried combining Stability AI's SDXL1.0 with AWS CLI.

Create a RAG application 1. AWS bedrock gen AI models are used to create embeddings and result generation. 2. FAISS library is used to create local pkl file of embeddings. 3. Streamlit is used for UI #RAG #GenerativeAI #AWSBedrock #FAISS

Here's a 3.5 minute video showing how to provision and run an LLM on RHEL AI in AWS. https://lnkd.in/eMTwyyDR Pretty easy!

Have you ever struggled with the Amazon Web Services (AWS)'s #awscdk documentary? Or perhaps you find CDK too verbose and you need a generate code while you simply type using your own language. CommandDash built a tool that indexes popular #github repositories. That means you can use chat based on the current codebase. Go try it out for AWS CDK and maybe you won't need their official docs anymore :D #ai #llm

These details help make AI real ! Great thread for someone trying to figure out AI infra.

This is where you rely on Hot Aisle Inc. to deploy and manage actual hardware for you, cut your costs and needs down significantly (no more hyperscaler tax), while controlling what you are running on and where your valuable data is stored. We are experts in this area and will work with you and Dell Technologies and Advizex to build solutions tailored to your needs. Let us do the heavy lifting of managing your capex and opex spend and get better value for your AI business.

Upfront investments in AI are huge. Every time new LLM comes out, scalers (cloud service providers) have to deploy new instances to run these models. Since existent apps already use previous generations of LLM, scalers can’t deprecate old versions easily. So, they have to buy new hardware. This induces demand for compute with unprecedented speed. And while returns for scalers are still in the future, HW manufacturers harvest profits today. With such pace, Nvidia will be worth $10T in less than a year.