Work Case study
Marjin AI
London. Partnership with Velocity.
Pulls scattered hospitality data into one place so people can ask questions in plain language.
The problem
A validated idea stuck in a Lovable prototype that tried to do too much and did not ship. Several earlier teams had not produced a product the founder could show buyers.
What shipped
ETL with API polling, webhooks and file polling. Bronze/silver/gold data; the app and AI query the gold layer only.
Role
One of the core engineers, not the sole builder, and client-facing. Velocity led product and design; One Eleven did the engineering. About 11 months. 336 commits. Owned the COGS editor and its edge tooling, fixed costs, the revenue-targets engine and evaluation cron, Deliverect/Zonal ingestion and canonical order refresh, and the cost/target tools the assistant (Maître) calls.