About the role
What You'll Work On
You will work on cutting edge computer vision and generative AI problems, particularly:
1. Virtual Try-On
• Develop AI models that can transform a person's photograph, jewellery, outfit, and background into realistic, high quality images.
• Work on improving image generation quality, consistency, and realism.
• Optimise models to make inference faster and more cost effective for large scale production.
2. Visual Similarity & Image Retrieval
• Build systems that can match a photograph of a jewellery piece to the exact product in a catalogue.
• Work with image embeddings, similarity search, and fine tuned vision models.
• Improve retrieval accuracy across visually similar products, designs, and variations.
Role Responsibilities
Research Scientist - Research & Innovation
• Own challenging, open-ended research problems in computer vision and generative AI.
• Study and build upon recent research papers in diffusion models, vision language models (VLMs), image generation, and embeddings.
• Design and conduct experiments to test new approaches and improve model performance.
• Develop novel solutions and translate promising research ideas into working models.
• Push the boundaries of image generation quality, visual understanding, and similarity matching.
Research Engineer - Model Development & Deployment
• Train, fine tune, evaluate, and optimise computer vision and generative AI models.
• Work with existing research architectures and adapt them to real-world product requirements.
• Build robust model training, inference, and evaluation pipelines.
• Deploy models into production and optimise them for speed, cost, and scalability.
• Collaborate with the research team to translate experimental findings into production-ready systems.
What We're Looking For
• Strong fundamentals and hands on expertise in computer vision, demonstrated through research papers, shipped products, or substantial technical projects.
• Experience with generative AI, diffusion models, image generation, vision language models, image embeddings, or visual similarity is highly relevant.
• Strong programming and model development skills, particularly in Python and deep learning frameworks such as PyTorch.
• Ability to independently identify problems, take ownership, experiment, and move quickly.
• A go-getter mindset with a strong bias towards execution and problem-solving.
• Startup experience and product sensibility - someone who cares not just about model performance but also about how the final output feels to the end user.
• Ability to document experiments, findings, iterations, and learnings clearly.
How We Work
• Clear, structured thinking approach problems sequentially and logically.
• Maintain notes and documentation at every stage of experimentation.
• Share frequent progress updates and learnings over Slack.
• Take ownership of problems and proactively drive them towards solutions.
• Work in a fast-paced environment where research needs to translate into tangible product outcomes.
Ideal Candidate Profile
Someone who enjoys solving difficult computer vision problems, is comfortable working with ambiguity, and can independently move from research and experimentation to real world implementation.
We are looking for people who combine strong technical depth with curiosity, speed, ownership, and an understanding of product quality.
This page is fully interactive when JavaScript is enabled. Please enable JavaScript to apply or browse related roles.
