[2023–2026]
FirstEdition/
Backend and admin for a collectibles market aggregator: collecting listings, matching them, AI enrichment, search. Contract work; the backend was a one-person job.
Laravel, PostgreSQL, Kafka, Flink, Elasticsearch, Redis$ stat firstedition
01_problemItem data arrives from many marketplaces in many shapes and has to become one catalogue.
02_howScrapers write to Kafka; streams are processed by Flink and Node processors; the catalogue and admin are Laravel; search is Elasticsearch.
03_whyStreams rather than nightly batches: marketplace data changes all the time.
04_result≈7,800 commits over 2023–2026.