Email QA Automation

Contract Forward Deployed Engineer · SiriusXM · July 2026–present. I work directly with the marketing team to understand operational problems and build internal applications around their workflows.

The problem

Marketing teams manually compare email proofs against campaign requirements, checking subject lines, preheaders, copy, CTA links, and tracking parameters. Line-by-line review is tedious and creates opportunities to miss mistakes.

What I built

I built an internal web application that compares email proofs with campaign requirements stored in Excel workbooks. It surfaces differences for marketers to review rather than deciding which version is correct.

How it works

• Load a campaign requirements workbook and add proofs in .eml, .msg, or HTML format. Multiple versions can be reviewed in one session.
• Compare subject lines, preheaders, body copy, CTA links, and tracking parameters.
• Choose a link-validation mode: Base validates and follows links; COI checks whether links are present without validating the URLs themselves.
• Review flagged discrepancies and use the change-log workflow to track the review.

Formats and integrations

Excel campaign requirement workbooks (.xlsx), HTML email proofs, and email files in .eml and .msg format.

Technologies used

• Front end: TypeScript, SolidJS, TanStack Router, Vite, and Tailwind CSS.
• Back end: Node.js and TypeScript APIs on AWS Lambda behind an Application Load Balancer.
• Data and AI infrastructure: DynamoDB and an OpenAI-compatible LiteLLM gateway.
• Development and delivery: Vite/Warp, a local DynamoDB mock, and GitHub Actions deployments through SiriusXM Builder/AWS.
• Outlook file parsing: @kenjiuno/msgreader for .msg proofs.

Value to the team

The application reduces manual line-by-line comparison and helps marketers identify potential errors before a campaign launches. Human reviewers retain the decision about which differences need correction.