From the course: Responsible AI on AWS: Bedrock Guardrails, Amazon Q Security, and SageMaker Clarify
Unlock this course with a free trial
Join today to access over 25,000 courses taught by industry experts.
Monitoring Bedrock calls
From the course: Responsible AI on AWS: Bedrock Guardrails, Amazon Q Security, and SageMaker Clarify
Monitoring Bedrock calls
- [Instructor] One problem with using foundation services like Bedrock is that you don't know exactly what's happening without looking at the logs. In this case, we actually can solve that problem by going to CloudTrail, which has default views into Amazon Bedrock. So we can look at all of the foundational model queries. You can see that the flow here is that first we have the event store and we have the query editor that allows us to communicate with that event store. Once we do a query, we say, you know, "Select star from Bedrock, et cetera." We're able to actually tabulate those results. So this is a good way to first get a interactive view without doing anything, setting up special logging or endpoints. You can actually just go right into CloudTrail and take a look at Bedrock calls. So let's go ahead and do that. I'm going to go over to CloudTrail here. Alright, so we're inside of CloudTrail, Lake, Query, and inside of here, one of the things that we can do is actually use English…
Contents
-
-
Course introduction1m 29s
-
(Locked)
AI security architecture4m 19s
-
(Locked)
AI auth patterns4m 1s
-
(Locked)
Complete AI security3m 49s
-
(Locked)
AI monitoring and logging3m 51s
-
(Locked)
AWS Rust compilation3m 49s
-
(Locked)
Monitoring Bedrock calls3m 46s
-
(Locked)
Visualizing Bedrock API calls2m 51s
-
-
-
-
-