I will build a computer vision app for video analytics counting and tracking
Applied AI Engineer building Computer Vision, RAG, AI Agents and Voice AI
About this Gig
Turn your cameras into useful data: people and vehicle counting, zone intrusion, queue monitoring, PPE checks, and fire, smoke or accident alerts.
I've built surveillance and safety systems (fire, accident and gas-leak detection) that run on live camera feeds, and retail computer vision used across 2M+ outlets.
What you get:
- YOLO detection with ByteTrack or DeepSORT tracking
- Counting lines, zones and dwell time on RTSP cameras or video files
- Alerts to Telegram, email or a webhook
- A simple dashboard or REST API (Premium), plus source code
I tune for your hardware so it runs in real time on a GPU, a regular PC or an edge device like Jetson.
Message me with a short sample clip and what you want to measure, and I'll confirm feasibility and the right package.
APIs:
Google Cloud Vision API
Programming language:
Python
Tools:
OpenCV
•
CVAT
•
PyTorch
Frameworks:
Keras
•
PyTorch
FAQ
Can it work with my existing CCTV cameras?
Yes, if they provide an RTSP stream (most IP cameras and NVRs do). Video files and webcams work too.
What hardware do I need?
It depends on the number of cameras and FPS. One or two streams often run on a regular PC; more streams need a GPU or Jetson. I'll recommend the cheapest setup that works.

