Architect & Digital Platform Lead
BBQ Help
An AI-powered barbecue assistant on web, iOS, and Android.
Visit site →
PythonTerraformDockerTypeScriptSwift
About this project
BBQHelp answers BBQ questions instantly (smoking times, temperatures, wood selection, recipes) as an AI assistant from the makers of Texas BBQ Rub. I architected the cloud infrastructure, application stack, and AI retrieval workflow behind it, and shipped it as a website plus native iOS and Android apps.
Technical specifications
- Backend
- Python 3.13, Flask (gunicorn/gevent), Docker
- Infrastructure
- AWS ECS Fargate + ALB, fully Terraform-managed
- AI
- AWS Bedrock Knowledge Bases (RAG + reranking), OpenAI API
- Web client
- React 19, TypeScript
- iOS client
- Native SwiftUI, iOS 17+, Swift 5, MVVM
- Mobile client
- Expo / React Native (SDK 54, RN 0.81), EAS Build
One backend serves three idiomatic, native-feeling clients: web, iOS, and Android. Each uses SSE streaming for a consistent real-time response experience. Supporting infrastructure (ECR, Lambda for scheduled jobs, DynamoDB for logging, CloudWatch, WAF, Route53/ACM) is entirely codified in Terraform.