Submitting a query

Before you look closely at how do Amazon Kendra responsoes look like, look at the results in the Kendra Search Console.

Try the query Which AWS service has 11 nines of durability?

Amazon Kendra’s response by default contains three types of results:

  • Relevant passages (It’s what in the search console is displayed as Kendra Suggested Answer)
  • Matching FAQs
  • Relevant documents

As you can see we have 2 “Kendra Suggested Answers”, 1 FAQ responses, and 69 total results.

Instead of looking at the query within the Kendra Search Console, look directly at the response Kendra returned to the Kendra Search Console.

On the Jupyter notebook or in your CLI with our Index ID, replace those variables and for the query use: Which AWS service has 11 nines of durability?

Using the example code you will receive this result:

Dive deep into the result using the following code:

response=kendra.query(
        QueryText = query,
        IndexId = index_id)
pp = pprint.PrettyPrinter(indent=1)
pp.pprint(response)

You will obtain a result like this: