Orfiq AI Solutions Ltd

We are looking for an ML Full-Stack Engineer (Computer Vision) to develop video analytics systems that evaluate non-technical skills in medical education. The focus is on simulation labs , where learners are assessed on AETCOM (Attitude, Ethics, and Communication) competencies — including empathy, ethical reasoning, and patient-centred communication. In this role, you will design and deploy AI pipelines that analyse simulation videos, detect behavioural and interaction patterns, and deliver actionable insights through dashboards. You will work across the stack — from video ingestion and multimodal AI models to backend APIs, front-end dashboards, and MLOps pipelines — ensuring solutions are scalable, explainable, and compliant with healthcare training standards. This role calls for a research and exploration mindset — someone who thrives on tackling open-ended problems, experimenting with emerging CV/ML methods, and turning abstract human behaviours into measurable, reliable AI insights. Responsibilities Design CV/ML pipelines to analyse learner–patient interactions in simulation labs, using YOLO, ViT, TimeSformer, Mediapipe, and Hugging Face vision models. Develop multimodal analytics by integrating video, audio, and transcripts to assess empathy, communication, and ethical decision-making. Ingest and process video data at scale using Azure Video Indexer and GPU-based compute environments. Build full-stack applications — backend APIs (FastAPI/Node.js) and front-end dashboards (React/Next.js) for educators. Ensure compliance and privacy , implementing features such as face blurring, de-identification, and data residency. Operationalize ML models with Azure ML pipelines, model registry, and CI/CD workflows. Collaborate with medical educators to align AI outputs with AETCOM rubrics and training objectives. Optimize compute resources and GPU usage through FinOps best practices. Nice to Have · Experience with WebRTC/live video streaming . · Background in medical education or CBME-based assessments. · Exposure to healthcare training, simulation-based education, or behavioural assessment . · Knowledge of multimodal learning (vision + audio + NLP) for behavioural analysis. · Expertise in FinOps for GPU cost optimization .