Principal AI Engineer

Tandem Search; Al Khobar; On-site; SAR 500,000 - 700,000; Full time

Job summary

Tandem Search seeks a Principal Computer Vision & AI Engineer to lead production AI systems for live video streams in real-world environments. This is a hands-on technical leadership role focused on architecture, implementation and deployment rather than people management.

You will guide senior engineers, own ML lifecycle from data strategy to monitoring, and optimize for edge and cloud inference while advancing multimodal/video-language models.

Qualifications

• Substantial hands-on experience with video-based computer vision systems.
• Production experience with live/streaming video.
• Proficiency in PyTorch and TensorFlow.
• Experience deploying AI models in production and edge devices.

Responsibilities

• Architect and build production computer vision and video analytics systems.
• Develop solutions across object detection, segmentation, tracking, and event understanding.
• Design and deploy VLM/multimodal AI systems for video.
• Own the full ML lifecycle: data strategy, experimentation, deployment, monitoring.
• Optimize models for edge and cloud inference focusing on latency and cost.
• Establish automated model evaluation, regression testing and production monitoring.
• Develop agentic AI capable of reasoning over video interacting with tools and cameras.
• Set technical architecture and standards while remaining hands-on.

Skills

Computer Vision

Video AI

PyTorch

TensorFlow

MLOps

Edge AI

Tools

YOLO

DETR/RT-DETR

GroundingDINO

VLMs (Qwen-VL/InternVL/LLaVA)

Job description

We are looking for a highly experienced Principal Computer Vision & AI Engineer to lead the technical direction of production AI systems operating over live video streams in real-world environments.

This is a hands-on technical leadership role rather than a people-management position. You will architect, build, optimise and deploy advanced computer vision, video understanding and multimodal/Vision-Language Model (VLM) systems while providing technical direction to senior engineers.

Key Responsibilities

• Architect and build production computer vision and video analytics systems.
• Develop solutions across object detection, segmentation, tracking, event understanding and spatial reasoning.
• Design and deploy VLM/multimodal AI systems operating over live and recorded video.
• Own the full ML lifecycle: data strategy, experimentation, fine-tuning, evaluation, deployment and monitoring.
• Optimise models for edge and cloud inference, balancing accuracy, latency, throughput and cost.
• Establish automated model evaluation, regression testing and production monitoring.
• Develop agentic AI systems capable of reasoning over video and interacting safely with tools, APIs, cameras and sensors.
• Set technical architecture and engineering standards while remaining highly hands-on with code.

What We're Looking For

• Deep hands-on experience in Computer Vision / Video AI.
• Strong production experience working with live or streaming video.
• Strong PyTorch and/or TensorFlow experience.
• Expertise across detection, segmentation and tracking using technologies such as YOLO, DETR/RT-DETR, GroundingDINO or equivalent.
• Production experience with Vision-Language Models / multimodal models such as Qwen-VL, InternVL, LLaVA, Gemini or equivalent.
• Experience deploying and optimising AI models in production.
• Exposure to edge AI, ideally TensorRT, ONNX Runtime, NVIDIA Jetson or similar.
• Strong understanding of model evaluation, MLOps, active learning and production monitoring.
• Ability to operate as a Principal/Staff-level technical authority while remaining hands-on.

Nice to Have

• Agentic AI / tool-calling systems such as LangGraph, LangChain or equivalent.
• Real-time CCTV, camera or video-streaming systems.
• Industrial, construction, robotics, autonomous systems or safety-critical AI experience.
• RAG / multimodal retrieval and vector search.
• Self-hosted VLM/LLM serving such as vLLM or NIM.

Critical requirement:

Candidates must have substantial hands-on experience with video-based computer vision systems. Pure GenAI/LLM profiles without strong production video/CV experience will not be suitable.