SecretSanta Matchmaker

This is a fun holiday project to help SecretSanta Matchmaking for extended families and friends. (After a heavy stretch of research and papers with serious sounding names (“... albation study of ...”), I felt it was time to bring some fun back into technology.)

🎄 What's Secret Santa?

• Its a gift-pairing tool that matches up people while respecting do's and do-not's.
• Think:
• ... Wolverine shouldn't be paired with Cyclops (relationship too cool),
• ... Peter Parker shouldn't be paired with Mary Jane (relationship too warm).

🔧 What was being explored?

• I was curious to explore the JAMstack idea a nudge further.
• This project blends the fun of svelte with the familiarity of a python backend.
• It also does away with microservices and manages it as a single monolothic deploy.

💻️ Keen to Try It?

• App Website - https://santa-service-mp4xuvosiq-as.a.run.app/static/ (its janky)

Stack

• Backend: Python, FastAPI, Uvicorn
• Frontend: Svelte
• Infrastructure: Docker -> Google Artifact Registry -> Google Cloud Run

Project Structure

• core - contains the core logic for the app
• frontend - contains the frontend code
• archive - contains the old code that was used to build the app
• tests - contains basic pytests for the app logic
• tests_bruno - contains the tests for the API endpoints

Other Notes

• Serving Static Assets - FastAPI used to serve the static assets from Svelte.
• Dev/Build: Makefiles used for most of the dev and build processes. Look through it for details
• Deploying to GCP: this was a real gotcha. in Dockerfile, make sure to specify the OS as linux/amd64. Otherwise, the build will fail with a cryptic error message.
• Bruno: a really simple and effective tool for testing APIs. Highly recommended for future projects.

Future Roadmaps

Features to add

• Admin Saver: Add a way to send out the assignments to the participants
• Value Add: add a way for participants to add their wishlists
• Engagement: Implement a bot-driven approach to participation, maybe in WA or a Telegram Chat
• Make it Interactive: Implement a graph visualization of the match-making process
• UI: make it MORE 90's CRT with frames or move to a more contemporary aesthetic.
• URL: link it as a subdomain to my personal website

Technologies / Approaches to try

• Think Different: Approach match-making as a graph problem, maybe using a graph database like Neo4J
• Make it TypeSafe: Use Typescript for the frontend
• Go "Cloudless": Use a CI/CD pipeline to deploy a custom VM on Digital Ocean or similar
• GraphQL: might as well try to implement this using GraphQL instead of REST