FinOps Dashboard Case Study

Cloud cost analysis often depends on manually maintained dashboards, specialized technical expertise, and hours of analyst effort. The FinOps Dashboard simplifies that process by allowing users to ask cloud cost questions in plain English and automatically generate current, interactive dashboards.

Dashboards generated in minutes instead of hours
Ask questions in plain English – no SQL expertise required
Insights refreshed on demand or on schedule
Self-healing queries with up to 5 automatic retries
Broader access to cloud cost intelligence across finance, engineering, and product teams

AWS environments continuously generate detailed cost and usage data. Turning that data into actionable insight, however, often requires SQL expertise, manual analysis, chart creation, and ongoing dashboard maintenance.

The solution connects plain-English questions directly to AWS billing data, automatically analyzing the data and creating the appropriate dashboard visualizations.

Example questions users can ask

What are our five most expensive AWS services?
Which months exceeded budget and by how much?
How has monthly cloud spending changed?
Which days exceeded a defined spending threshold?

Each question becomes a visualization within the published dashboard – no additional dashboard configuration required.

LambdaDashboardBuilder is a serverless AWS Lambda application that automates the workflow from question to published dashboard.

Natural Language to SQL: An Amazon Bedrock AI agent using Claude 3.5 Sonnet converts plain-English questions into Athena SQL queries using knowledge of AWS Cost and Usage Report (CUR) data, cost attributes, and budgets.
Automated Analysis: Queries run against AWS Cost and Usage Report data through Amazon Athena.
Self-Healing Queries: If a query fails, the error is returned to the AI agent, which corrects and retries the query up to five times.
Dynamic Visualization: AI selects the appropriate bar or pie chart and renders it using Google Charts.

Automated Publishing: Completed interactive HTML dashboards are published to Amazon S3 for stakeholder access.
Budget Awareness: Cloud spending can be compared against a configurable monthly budget.
Integrated Chat: Amazon Connect provides embedded chat for follow-up questions within the dashboard.
Flexible Invocation: The solution supports API Gateway, direct AWS Lambda invocation, or S3 prompt files.

Faster Cost Insights: Dashboards that previously required several hours of analyst effort can be generated in minutes.
Self-Service Access: Finance, engineering, product, and cloud teams can get the views they need without writing SQL.
Current Cost Visibility: Dashboards can be regenerated on demand or on a schedule using the latest billing data.

Reduced Manual Effort: Recurring report creation, dashboard maintenance, and query troubleshooting are significantly reduced.
Better-Informed Decisions: Faster visibility into spending patterns, budget performance, and cost drivers helps teams make more informed cloud cost decisions.

The Result: A traditionally manual and specialized reporting process becomes an AI-enabled, self-service FinOps capability that improves visibility, reduces effort, and accelerates cloud cost decisions.