Case study / 2025

Sentinel

Serverless email marketing platform with analytics, AI-assisted content generation, and multi-region AWS deployment.

sitecode

AWS Lambda, S3, API Gateway, DynamoDB, SQS, SNS, SES, Terraform

Sentinel interface

Problem

Email products need to coordinate content, audiences, event delivery, analytics, and failures. The project focused on serverless primitives rather than a monolithic backend.

System Shape

  1. 01Next.js dashboard manages campaigns and analytics views.
  2. 02API Gateway routes requests into Lambda functions.
  3. 03DynamoDB stores campaign and recipient state.
  4. 04SQS and SNS decouple asynchronous delivery workflows.
  5. 05SES handles outbound email while analytics events flow back into the dashboard.

Highlights

  • -Designed a multi-service AWS architecture with infrastructure-as-code.
  • -Used queues to isolate slow or failure-prone delivery work.
  • -Added real-time analytics and AI-assisted content generation.

Lessons

  • -Serverless systems still need explicit backpressure and retry thinking.
  • -The hard part is rarely one Lambda function; it is the contract between services.
  • -Operational visibility should be part of the product, not an afterthought.

What I Would Do Next

Add deeper observability around delivery failures, campaign-level retry policies, and cost profiling under load.