- frontend: SvelteKit & Vercel
- search engine: Algolia
- rest api & storage: Go & Firestore
- messaging service: Google Cloud Pub/Sub
- real-time analytics: BigQuery & CQRS architecture
- deployment: Docker/Docker-Compose & GCP & Traefik
- job board scraping Python
- why pub/sub?: previously was using RabbitMQ but we wanted more features (that consume from the same data) so for one-to-many messaging we made a switch to pub/sub
- why CQRS?: analytic queries could take a while so they should be calculated at write-time, also this keeps us in the 10tb query transfer data free tier of BigQuery
- why vercel?: original plan was to use Firebase, but Svelte 5 was hard to deploy on Firebase and Vercel is just very convenient
- why algolia?: no credit card required free plan
- why go?: front-facing API is a serverless function, so the quick cold starts were useful. also, Algolia and BigQuery consumers do potentially long background processing, so the strong concurrency model was critical for scale
- why firestore?: speed is of upmost importance... it also has a free tier
- why traefik?: automatically handles SSL certification renewal which Nginx doesn't natively handle and does not support hot renewal with new certificates
