GiM / GiMOps
Hospitality operations intelligence shaped around legacy venue infrastructure.
- Operational discovery
- Mapped recurring venue problems across old hardware, weak Wi-Fi, spreadsheets, staff memory, and close-down workflows.
- Implementation approach
- Translated those constraints into a read-only, low-friction product direction with clear privacy boundaries and useful tools before reporting.
- What I built
- Public multi-page site, product and trust documentation, Cloudflare Worker contact flow, and internal Python/FastAPI/SQLite signal prototypes.
- Evidence
- Live gimindex.com site, validated contact API, plus an internal SVP edge encoder, test report, and edge-to-cloud demo stack.
Semantic Vision Protocol
SVP is protocol and prototype work for moving GiM beyond isolated edge deployments. In the prototype, local processing supported privacy but had to work within limited hardware and RAM; coordinating multiple venues required a practical path into cloud infrastructure.
Instead of transferring raw video with its privacy, bandwidth, and storage costs, SVP converts locally processed observations into small, anonymised semantic records before cloud transfer. This preserves useful operational signals without transferring identifiable raw footage.
- Shows requirements discovery, privacy-conscious architecture, edge-to-cloud implementation thinking, infrastructure constraints, and multi-site design.
- Status: active project with public site; SVP remains protocol and prototype work, not a production deployment.
