Managed RAG Deployment on Amazon Bedrock - Deployed in Minutes
Вставка
- Опубліковано 11 гру 2024
- This video shows you how to implement a fully managed RAG deployment on Amazon Bedrock using Anthropic Claude models and AWS Python SDKs
▬▬▬▬▬▬ Connect with me 🔗 ▬▬▬▬▬▬
LINKEDIN ► / trevspires
TWITTER ► / trevspires
This is great! I just went through this the other day and was able to RAG all the transcripts from my livestreams...
Thanks for the fire video homie! It's super simple and easy to follow.
Subbed. Amazing video and awesome that we can replace about 20 startups with something we can spin up in minutes.
This helped me understand RAG in the context of Amazon Bedrock really well. Thanks for the clear and detailed explanation!
What? Not a 30 min video with 25 being worthless filler?
Earned a sub, u a baller
Your video was very helpful, it gave me a good idea of how to start estimating costs to propose a project for the company (annual revenue of +10B) I work for
Super simple here!. Thanks Buddy.
thank you！ great content
What set-up do you need before these steps, I am having difficulties and "no such index [bedrock-knowledge-base-default-index]"
great stuff broski, any chance you could do something similar with Sagemaker?
What you looking for on Sagemaker?
Maybe I can. I’d like an excuse to spend more time on it
Thank you! How about embedding PDF files and stored in S3 ？
excellent. thank you
right, but how do we get this to function in vs code? :)
Hi, Trevor, i' m looking for a for a software to build some flowchart like yours. Please could you give me the name. Thanks
subscribed, appreciate it!keep it up!
I have an application that stores specific documents by folder identified by a unique ID, then I process them with Textract and save the information in another S3 following the same hierarchy. Can I create a chatbot that uses the information from a specific document during the user's session?
Yes? But I'm not sure on the details for your implementation here
@@trevspiresI create the context with the information from a specific document, in this case documents with information about sales quotes, and I create the chatbot with that context, so that the user can get the information they need from the document without having to read it completely
Awesome video, Trevor. Thank you! What tools and architecture would you recommend if I wanted to build a RAG app like this but for multiple users and I needed to keep their documents separate and private during retrieval? Thanks again! (Subscribed)
Check out Amazon Q!! It has some good RBAC capabilities built in top of IAM.
You could also explore using "Metadata Filters" for RBAC within RAG. This workshop has some good content here for it - catalog.workshops.aws/persona-based-access-genai-application/en-US/01-introduction
Awesome video. Can you share the pricing structure for that basic set up? If you use Pinecone (free tier) is there no cost whatsoever?
I’m not a pinecone pricing expert, not sure how long they’ll have a free option
You can also run a local VectorDB
What type of files can I use for my knowledge base? If I have a pdf file, will I need a script to convert it to text beforehand?
hello sir!!! I have a issue in creating the knowledge base. When I create it shows failed to create open search serverless collection.eventhough I gave the full access access for bedrock and opensearch service for the user and made the s3 bucket to be accessed by the opensearch service, the issues is not fixed. can you help me to clear that issue? I'm struggling with that issue !!! please help me !!!
I also got this issue. Did you by any chance find a solution? Or could you maybe help @trevspires?
awesome video, quick question, can it parse PDF documents as opposed to text docs in S3 bucket?
Yep!!! Supports most unstructured and semi-structured docs.
Specific supported types should be in the Bedrock KB docs
@@trevspires Thank you! Also, is there a way to use LLaMA for the model? Right now, I got access to all the models but I can only use Titan for RAG
Okay, but I thought you were going to show the step by step for the connection. And I didn't find a series of videos or the following video where it is explained. By any chance, does this exist? or was this intro nothing more? Thankssss
I don’t have a video, but there are knowledge base workshops from AWS you can check out
Question: What if I ask a random question which isnt in the document, what will I be expecting? What if I asked, who is trevor?
Good stuff. Thank you!
Awesome video! How much money did this test cost to you?
Less than $2
unable to ACCESS ANTHROPIC MODEL ACCESS in aws ..
You turn on model access?