2 months ago
London, United KingdomSenior
Responsibilities
- Design and build generative AI applications for complex financial and legal workflows.
- Lead end-to-end model development, including reproducible research, delivery, and deployment best practices.
- Evaluate, adapt, fine-tune, and deploy open-weight language models for production use cases.
- Build high-quality and synthetic training datasets and data flywheels to improve model quality.
- Develop rigorous benchmarks, evaluation frameworks, and regression testing systems.
- Build and operate training pipelines, deployment systems, evaluation infrastructure, and production model monitoring.
- Apply distributed training, parameter-efficient fine-tuning, GPU optimization, inference acceleration, distillation, and quantization.
- Collaborate cross-functionally, mentor teammates, communicate ideas, and promote AI adoption across the organization.
Requirements
- Deep experience evaluating, adapting, and deploying open-weight models such as Llama, Qwen, Mistral, and Gemma.
- Expertise in supervised fine-tuning, preference optimization, instruction tuning, and domain adaptation.
- Strong experience with training data, synthetic data generation, model evaluation, and production ML engineering.
- Experience applying AI to financial and legal domains such as credit, restructuring, debt instruments, filings, court documents, and capital markets workflows.
- Familiarity with distributed training, parameter-efficient fine-tuning, GPU optimization, inference acceleration, and model optimization techniques.
- Strong proficiency in Python for production-ready delivery.
- Strong product focus and interest in understanding end-to-end impact for users.
- Good knowledge of AWS AI infrastructure.
Benefits
- Competitive salary, equity, pension with 9fin matching up to 7%, private medical insurance, income protection, group life assurance, season ticket loan, and Cycle to Work scheme.
- 25 holiday days per year, local public holidays, enhanced parental leave, and flexible working arrangements.
- Hybrid working model with the option to work abroad for up to 3 months per year.
- One month of paid sabbatical after 5 years of service.
- Professional learning and development budget, an £800 UK AI experimentation budget, and company social events.
Categories
