GenAI Data Assistant for HR Analytics

EFS DevOps built a GenAI-powered self-service analytics assistant that translates natural language into SQL queries — reducing report creation from 2 days to 15 minutes and achieving 4.5x ROI within 6 months.

About the Customer

Industry: Enterprise Data & Operations

Managers lacked technical skills to query Amazon Athena or QuickSight directly, causing delays and constant ad-hoc report requests. The analytics team was overwhelmed with routine data pulls, leaving no capacity for strategic analysis.

The Challenge

• 150 ad-hoc report requests per month from business managers
• 2-day average per report due to data complexity and manual analysis
• 8-person analytics team consumed by 100% tactical reporting
• No self-service capability for non-technical users
• $5.20 cost per report in analyst labor

AWS Solution Architecture

• Amazon Bedrock (Claude 3 Instant) — Natural-language to SQL translation
• Amazon Athena + Glue Catalog — Query execution and schema management
• OpenSearch Serverless + Titan Embeddings — Semantic context and metadata grounding
• QuickSight API — Automatic visual summary rendering
• Amazon Verified Permissions — Row-level security enforcement per user role
• SageMaker Pipeline — Continuous learning through user feedback on query accuracy

Fully serverless with $0 idle cost. Per-query cost: ~$0.002.

Results

4.5x ROI within 6 months. Line managers now self-serve data insights. Analysts reallocated to strategic projects.

AWS Services

Amazon Bedrock (Claude 3 Instant), Amazon Athena, AWS Glue Catalog, OpenSearch Serverless, Amazon Titan Embeddings, Amazon QuickSight, Amazon Verified Permissions, SageMaker Pipeline, AWS Step Functions, AWS Lambda.