Quality assurance · Production

BQA streamlines education quality reporting using Amazon Bedrock and generative AI

The problem

BQA's manual review of self-evaluation reports was slow and error-prone: institutions submitted incomplete or inaccurate information, supporting evidence frequently failed to substantiate claims, and staff spent significant time and resources following up with institutions to rectify submissions.

Workflow diagram · grounded in source
1
Document upload to S3
trigger
“Relevant documents are uploaded and stored in an Amazon Simple Storage Service (Amazon S3) bucket.”
2
SQS event notification
integration
“An event notification is sent to an Amazon Simple Queue Service (Amazon SQS) queue to align each file for further processing. Amazon SQS serves as a buffer, enabling the different components to send and receive messages in a reliable man…”
3
Text extraction via Textract
ai_action
“The text extraction AWS Lambda function is invoked by the SQS queue, processing each queued file and using Amazon Textract to extract text from the documents.”
4
Summarization via Meta Llama
ai_action
“The text summarization Lambda function is invoked by this new queue containing the extracted text. This function sends a request to SageMaker JumpStart, where a Meta Llama text generation model is deployed to summarize the content based …”
5
Compliance assessment via SageMaker
validation
“In parallel, the InvokeSageMaker Lambda function is invoked to perform comparisons and assessments. It compares the extracted text against the BQA standards that the model was trained on, evaluating the text for compliance, quality, and …”
6
Results stored in DynamoDB
integration
“The summarized data and assessment results are stored in an Amazon DynamoDB table”
7
Bedrock AI evaluation generation
ai_action
“the InvokeBedrock Lambda function invokes Amazon Bedrock to generate generative AI summaries and comments. The function constructs a detailed prompt designed to guide the Amazon Titan Express model in evaluating the university's submission.”
8
Scored evaluation output
output
“Assign a score from 1–5 for each comment, citing evidence directly from the content”
Reported outcome

BQA's proof of concept with Amazon Bedrock anticipates faster turnaround times for generating 70% accurate and standards-compliant reports, a 30% reduction in evidence analysis time, and 30% reduced operational costs through process optimizations.

Reported metrics
Report accuracy (anticipated)70%
Evidence analysis time reduction (anticipated)30%
Operational cost reduction (anticipated)30%
Reported stack
Amazon BedrockAmazon SageMaker JumpStartMeta Llama
Source
https://aws.amazon.com/blogs/machine-learning/how-bqa-streamlines-education-quality-reporting-using-amazon-bedrock?tag=soumet-20
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

BQA's proof of concept with Amazon Bedrock anticipates faster turnaround times for generating 70% accurate and standards-compliant reports, a 30% reduction in evidence analysis time, and 30% reduced operational costs…

What tools did this team use?

Amazon Bedrock, Amazon SageMaker JumpStart, Meta Llama.

What results were reported?

Report accuracy (anticipated): 70%; Evidence analysis time reduction (anticipated): 30%; Operational cost reduction (anticipated): 30% (source-reported, not independently verified).

How is this quality assurance AI workflow structured?

Document upload to S3 → SQS event notification → Text extraction via Textract → Summarization via Meta Llama → Compliance assessment via SageMaker → Results stored in DynamoDB → Bedrock AI evaluation generation → Scored evaluation output.

Related quality assurance cases