Inspiration

Business Analysts spend hours turning ideas into Jira tickets and Confluence documentation. These repetitive tasks slow down delivery and distract from strategic thinking.
We wanted to create an autonomous assistant that understands what a BA wants, plans its actions, and executes them automatically.
That idea became AI BA Agent - our "digital business analyst" built entirely on AWS.


How We Built It

We built a serverless architecture on AWS using:

  • AWS Lambda (Python 3.11), API Gateway, and S3 for logic and storage
  • Amazon Bedrock (Nova Pro) for reasoning and decision-making
  • Next.js, TypeScript, and ShadCN UI for a clean, responsive frontend
  • Secure integrations with Atlassian Jira Cloud API and Confluence API for automatic creation of epics, stories, and documentation
  • CloudFront, WAF, IAM, and CloudWatch for security, delivery, and monitoring

What We Learned

  • How to orchestrate autonomous AI workflows using AWS Bedrock and Lambda
  • Designing multi-step reasoning chains for agentic behavior
  • Securely managing secrets and API keys in AWS IAM and Secrets Manager
  • Structuring prompts and validation logic for consistent enterprise-grade outputs

Despite these challenges, we delivered a working autonomous agent that turns natural-language business requests into real Jira tickets and Confluence pages - all powered by AWS Bedrock.

Built With

Share this project:

Updates