|
Full-Stack Developer focused on building and shipping production web applications. Built a QR-based food ordering platform that was discovered on GitHub and led to a $1,200 client engagement to scale it into a full Dining OS with real-time table management and Stripe Connect integration. I work across frontend, backend, cloud infrastructure, and DevOps, with hands-on experience building RAG and vector-search systems.
|
|
Additional: TanStack Query · Zustand · Axios · Stripe SDK · shadcn/ui · Swagger · Supertest · Trivy · GitFlow · Qdrant · LangChain · RAG · embeddings · indexing · re-ranking
| Project | Description |
|---|---|
| Complex : Food Court Order Management System | Multi-tenant SaaS that lets diners scan a QR code at the table and order from any restaurant in a food court using one cart, with no app install or queuing.Next.js Express Prisma Stripe Connect PostgreSQL |
| Coursera Multimodal Intelligence Platform | Internal platform for course teams with ingestion and RAG pipelines. Course assets are chunked, embedded and indexed in Qdrant, then queried through FastAPI with re-ranking and LLM reasoning.Next.js TypeScript FastAPI Qdrant LangChain TanStack Query shadcn/ui |
| Cloud-Native CI/CD & Infrastructure on Azure | Production-style cloud infrastructure provisioned with Terraform, including Azure networking, ACR and AKS with managed identities. GitHub Actions automates testing, builds, Docker packaging and AKS deployment with health checks for zero-downtime rollouts.Terraform Azure AKS ACR Docker Kubernetes GitHub Actions |
| Fastmenu : Dining OS Platform | Production dining platform evolved from the open-source Dine Easy QR ordering MVP. Expanded with role-based access control, Stripe Connect for payments and vendor onboarding, plus Google APIs for translation and location features.React Redux Material UI Firebase Stripe Connect Google APIs |
- Delivered a $1,200 freelance client product as a sole developer.
- Expanded a QR ordering MVP into a full Dining OS over an 11-month engagement.
- Coordinated a 9-person project team across frontend, backend, database and RAG workstreams.
- Reduced a production Docker image by over 90% to 25 MB using multi-stage builds.
- Built a VBA reporting tool that eliminated approximately 2 hours of manual reporting per day across three departments.