1st place, advanced category — built and shipped in a weekend

Led a team of 3 to first place with a real-time resident face-identification system for assisted-living safety monitoring.

Role

Team lead (3 people) — architecture, CV pipeline, delivery

Goal

Real-time resident identification for assisted-living safety

Key outcome

1st place, advanced category

Artifact

secure-holmes on GitHub ↗

1st

place — advanced category, judged by Voxel51

person team, led end-to-end under hackathon pressure

48h

from idea to working real-time demo

$ cat problem.md

Who's at the door — and should they be?

Assisted-living facilities need to know when residents wander into unsafe areas or unknown people enter — in real time, from live video, with an interface staff can actually use.

$ ./approach.sh

01Built real-time face identification with DeepFace over live video for resident recognition.

02Shipped a React frontend so the system was a usable product, not a script.

03Led the team of 3 — split the work, integrated continuously, demoed live.

$ cat results.json

1st place, advanced category at Voxel51's Visual AI Hackathon
Working real-time demo presented live to judges
Code public — secure-holmes on GitHub

$ grep -r skills

rapid prototypingteam leadershipDeepFace · face recognitionreal-time videoReactlive demo delivery