Project Description
Project: Pixie Smart Camera + App Prototype
Objective:
We are developing a smart camera system (Pixie) that behaves like a virtual camera operator.
We are looking for a developer (or small team) to:
1. Build and refine the React Native Android application
2. Integrate real-time video processing (tracking + auto zoom)
3. Assist in sourcing or building a suitable prototype camera system
This prototype will be used for investor demonstrations and product validation.
We already have working prototype logic that:
Performs real-time subject tracking
Applies smooth, natural auto zoom (PID-style motion)
Seamlessly transitions between face tracking and body tracking
Your role is to bring this to life in a stable, real-time system.
Core Requirements
1. Video Pipeline:
* Input: RTSP stream from Wi-Fi camera
* Real-time decoding and frame processing on Android
* Low-latency performance (target <200ms)
Notes:
* Processing can be done at 1080p / 2.7K
* 4K is desirable for output quality, not required for processing
2. Application Development (React Native + Native Modules)
You will:
* Integrate RTSP video into the app
* Pass frames to processing layer (native if needed)
* Render processed output (tracking + auto zoom)
* Build a simple but functional UI for testing.
Important:
You may need to use native Android (Java/Kotlin/C++) for performance-critical parts
3. Tracking & Auto Zoom Integration:
We already have:
* Working tracking + auto zoom logic (prototype stage)
You will:
* Integrate and optimise this for real-time performance
* Ensure smooth, stable behaviour
* Maintain low latency
4. Connectivity Architecture (IMPORTANT):
We are not using AP-mode cameras.
Required setup:
* Camera operates in Wi-Fi STA mode
* Camera connects to:
* either a router OR preferably a tablet hotspot (self-contained system)
* App and camera communicate over local network (LAN)
Streaming:
* RTSP (primary)
* WebRTC optional (bonus)
5. Camera (NEW REQUIREMENT):
* We are currently sourcing the correct camera hardware.
* We are looking for someone who can also assist with one of the following:
Option A (Preferred):
* Recommend and source a suitable off-the-shelf camera with:
* STA mode
* RTSP stream
* Low latency
Option B:
* Build a prototype camera system, for example:
* embedded Linux (e.g. Raspberry Pi or similar)
* custom RTSP/WebRTC pipeline
This is not mandatory, but a strong advantage
6. Power Requirements:
* Battery-powered preferred
* USB-C charging
* For prototype: external battery acceptable
Deliverables:
* Working Android app (React Native + native components)
* RTSP video successfully integrated
* Real-time tracking + auto zoom functioning
* Stable low-latency performance
* Camera integrated and working with the system
Basic documentation:
* setup instructions
* system architecture
Success Criteria:
* Smooth, natural tracking behaviour
* Stable auto zoom
* Low-latency live video
* Reliable connection between camera and app
* System suitable for investor demo
Important Notes:
* We do not currently have a final camera solution
* Previous camera tested (AP mode) is not suitable
* Flexibility and problem-solving are critical
Ideal Candidate:
* Strong experience with:
* React Native
* Android native development
* real-time video (RTSP/WebRTC)
Experience with:
* OpenCV / computer vision
* streaming pipelines
* embedded systems (bonus)
Final Note:
We are not looking for a generic app developer.
We are building a real-time, hardware-integrated product.
If you can contribute across both software and hardware (even at a prototype level), that is a major advantage.