Computer Vision & AI Lead

  ·  Al Khobar, KSA  ·  Full-time  ·  Hands-on technical leadership (player-coach)

Millions of people work on large industrial and construction sites where a missed hazard can cost a life. You'll lead the AI that turns live site video into real-time safety insight — building perception systems trusted to make the right call, at scale, in the field. This is hands-on technical leadership: you architect and build, and you set the direction for a small, senior AI team.

You'll lead a small team of senior ML/CV engineers and own our computer-vision safety product end to end — from perception models through evaluation and deployment, across both edge and cloud. You make the architectural calls, stay hands-on in the code, and are the person accountable for the accuracy, reliability, and direction of the AI.

Set the technical direction for our perception stack — deciding where modern vision-language and multimodal models add the most value, and where lighter, purpose-built models fit.Build agentic AI systems that reason over live video, make decisions, and operate in a closed loop with real-world cameras and sensors.Own the full model lifecycle: problem framing, data, fine-tuning, deployment, and monitoring.Design the evaluation methodology that judges model quality — task-appropriate metrics and evaluation sets for both detectors and open-ended vision-language / agent outputs (precision/recall, calibration, human-agreement, LLM-as-judge, faithfulness checks), with regression gates so no model, prompt, or agent change ships without a measured quality bar.Run a continuous data flywheel — use active learning (uncertainty and edge-case sampling) and production review signals to focus labeling where it most improves the model, and to sustain quality as site conditions change.Deploy across edge and cloud under real latency, cost, and connectivity constraints, with consistent results either way.Build responsible-AI guardrails for a safety-critical domain: confidence thresholds, human-in-the-loop review for high-severity events, and audit trails.Lead and mentor the team — hiring, technical growth, code quality, and delivery in partnership with Product.

~7–10 years building and shipping software, including 5+ years in ML / computer vision in production (PyTorch or TensorFlow) across tasks such as object detection, segmentation, and tracking.Hands-on experience with vision-language / multimodal models (open-weight and API-based) — prompting, grounding, and applying them to real visual-understanding problems.Experience designing agentic, tool-using AI systems — LLM-driven agents that reason, call tools/APIs, and take actions autonomously.Model adaptation and fine-tuning (supervised and parameter-efficient methods such as LoRA), including dataset curation and active-learning workflows that target labeling efficiently.Proven deployment across edge and cloud, optimizing inference for latency, cost, and throughput on real-time targets.Deep model-evaluation and MLOps experience — designing metrics and evaluation sets for detectors and for open-ended multimodal / agent outputs (precision/recall, calibration, human-agreement, LLM-as-judge), with regression and A/B testing, experiment tracking, versioning, and production monitoring / observability for drift and quality.2+ years leading or tech-leading engineers, with strong Python and software-engineering fundamentals.

Retrieval-augmented generation (RAG) and vector-search / retrieval pipelines.Real-time video / streaming-camera systems, or agents that operate physical devices via standard control protocols.Edge-AI toolchains and accelerators (e.g., TensorRT, ONNX Runtime, on-device GPU).Experience in safety-critical, industrial, or other real-world operational environments.Open-source contributions or publications in top computer-vision / ML venues.

Real-world impact — your work helps keep people safe on some of the world’s most demanding sites. Production systems, not pilots.Ownership and autonomy — you architect and own the full lifecycle; your name is on the architecture.Modern stack and real data — current foundation-model tooling, real site data, and edge + cloud compute to experiment with.Growth — a clear path toward Staff / Principal / Architect as the team scales.

Competitive salary (basic + housing + transport), performance bonus, and equity participation.Relocation and visa sponsorship for candidates joining from outside KSA.Health insurance, annual flights, and standard benefits.Based in Al Khobar, KSA. Excellent written and spoken English required; Arabic a plus.