Join us to build the future of AI-Powered
Empower intelligent cameras, real-time analytics, and autonomous perception everywhere.
We’re building the AI video analytics platform that enables machines to see, understand, and act in real time — from edge devices to massive distributed systems. Our technology powers smart cities, retail insights, factory automation, and intelligent security through an end-to-end software stack: optimized deep learning inference, hardware acceleration (GPU/NPU), and flexible cloud-edge orchestration.
From our custom AI Processor API to modular video pipelines, every part of CVEDIXGTR is engineered for scalability and performance. We develop our own SDKs, build on top of open standards like GStreamer and RKNN, and integrate seamlessly with NVIDIA, Rockchip, and OpenVINO ecosystems.
Our mission is simple: make vision AI fast, open, and everywhere, any-hardware.
No hype — just relentless engineering, iteration, and shipping production-ready intelligence at the edge.
No hype, no fluff, MBAs — only deep tech, real data, and results that run.
Company: CVEDIX
Location: Vietnam (or Remote)
Type: Full-time/Part-time/Internship
Salary: 0 - 1000$
Job Description
- Design and implement intelligent vision systems that enhance real-time video analytics and edge AI inference performance.
You will work on the CVEDIX - General Tensor Runtime Platform, developing scalable machine learning pipelines that process multi-camera video streams across GPUs, NPUs, VPUs and heterogeneous compute environments.
- Your primary mission is to bridge research and production — transforming deep learning models into deployable, efficient, and reliable AI services running on edge devices and smart infrastructure.
Responsibilities
- Design, train, and optimize deep learning models for object detection, face recognition, tracking, and behavior analytics, etc.
- Build robust validation and benchmarking pipelines using real-world video data and synthetic simulation tools.
- Develop automated testing, continuous training (MLOps), and monitoring pipelines for deployed AI modules.
- Collaborate with the Video Processor, AI Processor, Core Runtime, and Output Processor teams to integrate ML components into CVEDIXGTR’s modular C++/Python APIs.
- Improve inference performance through quantization, pruning, INT8 calibration, and multi-backend deployment (RKNN, TensorRT, OpenVINO, Qualcomm, etc).
- Build and maintain CI/CD workflows for model deployment, regression testing, and on-device benchmarking.
Preferred Skills
- Strong proficiency in Python, C++, and ML frameworks such as PyTorch / ONNX / TensorFlow.
- Experience with edge inference SDKs (RKNN, TensorRT, OpenVINO, Qualcomm, etc) and hardware acceleration (GPU/NPU).
- Familiarity with GStreamer, OpenCV, and video pipeline optimization.
- Knowledge of Docker, CI/CD, API design, and MLOps automation.
- Background in computer vision, robotics, automations or embedded AI systems.
What We Value
- No hype. No bureaucracy. No corporate noise.
- Just engineers solving real-world vision problems and shipping production-ready AI that runs everywhere, anywhere.
Compensation
- Competitive salary based on experience and performance, with opportunities for equity participation in the CVEDIX ecosystem.
Who we're looking for
- We don't care where, or even if, you went to school. We don't care if you have a traditional background. We're just looking for people who can independently contribute to shipping my products.
- We love a great GitHub, open source contributors, good projects, and competition winners.
Contact
- Interested applicants may send their portfolio or GitHub/LinkedIn to
- 📧 hr@cvedix.com
- Subject: Machine/Deep Learning Engineer – CVEDIX