£30 Per hour
Inside
Hybrid
London Area, United Kingdom
Summary: The Data Analyst role focuses on evaluating generative AI models, specifically in image and video generation, for a leading social media brand. The position involves collaborating with researchers to design evaluation processes, conducting manual assessments, and identifying performance gaps. The role is contract-based for 12 months and requires a strong attention to detail and foundational knowledge in AI concepts. The position is primarily on-site for three days a week in London.
Key Responsibilities:
- Collaborate with researchers and engineers to design and prototype human rating and semi-automated evaluation processes for generative AI models
- Conduct in-house manual evaluations of AI-generated images and videos particularly for complex or high-impact tasks
- Identify and summarise model performance gaps, using both qualitative and data-driven approaches
- Support the creation of training datasets by mining data that targets specific model weaknesses
- Help scale validated evaluation protocols in partnership with production and data operations teams including process coordination and auditing
- Contribute to iterative development cycles that feed directly into model improvements
Key Skills:
- Strong attention to detail and a mindset for experimental problem-solving
- Clear written and verbal communication skills
- Basic working knowledge of Python, Unix, and SQL
- Ability to manage time independently while collaborating effectively with a team
- A solid foundation in computer vision and/or generative AI concepts
Salary (Rate): £30.00/hr
City: London
Country: United Kingdom
Working Arrangements: hybrid
IR35 Status: inside IR35
Seniority Level: undetermined
Industry: IT
Data Analyst – Generative AI Evaluation (Image & Video) Contract 12 months Location: London (3 days onsite) £250 PD – Inside IR35 One of the worlds leading social media brands is looking for Data Analysts to join a pioneering AI research team working on image and video generation. From converting text into images to transforming images into video, this team is pushing the boundaries of generative AI. Your work will directly contribute to improving how model quality is measured and scaled
Role and Responsibilities
- Collaborate with researchers and engineers to design and prototype human rating and semi-automated evaluation processes for generative AI models
- Conduct in-house manual evaluations of AI-generated images and videos particularly for complex or high-impact tasks
- Identify and summarise model performance gaps, using both qualitative and data-driven approaches
- Support the creation of training datasets by mining data that targets specific model weaknesses
- Help scale validated evaluation protocols in partnership with production and data operations teams including process coordination and auditing
- Contribute to iterative development cycles that feed directly into model improvements
Skills and Experience
- Strong attention to detail and a mindset for experimental problem-solving
- Clear written and verbal communication skills
- Basic working knowledge of Python, Unix, and SQL
- Ability to manage time independently while collaborating effectively with a team
- A solid foundation in computer vision and/or generative AI concepts