As Bedrock unveiled a few weeks ago, there is still a lot of turmoil in the domain, and I felt that my personal AI assistant, created earlier this year, had to keep up the pace by using its own data as part of its “knowledge”. Let’s recap the journey.
The first article presented the AI assistant based on OpenAI ChatGpt.
Build your personal speaker assistant with Amplify and ChatGPT
As a speaker, it can be difficult to keep track of all the information you need to convey during a presentation. In…
As soon as Bedrock was generally available, I re-purposed the assistant to utilize Bedrock A21Labs’ Jurassic Ultra foundation model to achieve similar behaviour.
Bedrock unveiled: A Quick Lambda example
This article, instead, outlines how to use the Bedrock Embedding Titan model to index documents towards an OpenSearch Serverless vector database.
The ultimate goal, outlined in the next article, is to use such indexed documents and more information (chat memories and Bedrock FM) so that the AI assistant provides more contextual answers.