Data Scientist – Computer Vision & Generative AI

About the role

We are looking for a Data Scientist specialized in Computer Vision and Generative AI to join our AI team. The role involves developing, fine‑tuning and deploying state‑of‑the‑art vision and diffusion models for real‑world applications. You will collaborate with engineering to bring advanced image understanding and generation solutions into production.

Key responsibilities

• Design, train, fine‑tune and deploy computer vision and generative AI models.
• Develop solutions for object detection, segmentation, depth estimation, image inpainting and virtual staging.
• Build and optimise end‑to‑end pipelines for image understanding and generation tasks.
• Evaluate model performance with appropriate metrics and implement improvements.
• Create and maintain data annotation, training, validation and testing workflows.
• Work closely with engineering teams to productionise AI models and services.
• Research and implement the latest advancements in computer vision, diffusion models and multimodal AI systems.
• Optimise models for inference speed, memory consumption and scalability.
• Develop robust APIs and model‑serving solutions for production environments.
• Document experiments, model architectures and deployment processes.

Required profile

• 3+ years of hands‑on experience in Machine Learning, Deep Learning, Computer Vision and Generative AI.
• Proven experience developing, optimising and deploying production‑grade AI solutions.
• Strong expertise with vision models such as RF‑DETR, DETR variants, YOLO family, Faster R‑CNN, Mask2Former, Segment Anything Model, semantic/instance segmentation and depth estimation models.
• Hands‑on experience with diffusion models like Stable Diffusion XL, ControlNet, image inpainting/outpainting, LoRA training and Hugging Face Diffusers.
• Solid understanding of CNNs, Transformers, Vision Transformers and modern deep‑learning architectures.

Required skills

• Python programming
• PyTorch
• FastAPI or Flask
• Docker and containerised deployments
• Linux environments
• Git version control
• Cloud platforms: AWS, GCP, Azure or RunPod
• Data annotation tools: CVAT, Label Studio, Roboflow
• Hugging Face Diffusers
• Stable Diffusion XL, ControlNet, LoRA
• RF‑DETR, DETR, YOLO, Faster R‑CNN, Mask2Former, Segment Anything Model
• Depth Anything, Vision Transformers, CNNs, Transformers