Are agents the missing piece for LLMs to super-charge real-world adoption? Today, I had the opportunity to attend AWS Immersion Day at AWS Helsinki office and try hands-on workshop on Agents, RAG, and more.
Thanks Amazon Web Services (AWS), Ankit Nigam, Kimmo Isosomppi, Heidi Tähtinen, Ezaldeen Arafat for organizing this workshop. This was immensely useful and improved my understanding of Agents, RAG, Guardrails and inference.
As part of the hands-on workshop, I have created a mortage-processing-agent which uses RAG to add domain specific knowledge base and then created agents which can determine if end-user can afford it based on their financial situation.
After that, created Amazon Bedrock Flows that orchestrates entire workflow. Last but not the least, created Guardrails to implement safeguards and enforce responsible AI policies for my GenAI application.
Here is the mortgage processing agent I built during the hands-on workshop.
The highlighted yellow box is the prompt user enters followed by the response from Bedrock.
All in all, this was a very good way to get hands-on experience on Amazon Bedrock Agents, RAG, Flows and Guardrails.
Top comments (0)