LLM Engineer
Ultralytics4 months ago
London, United KingdomSenior
Responsibilities
- Develop and scale LLM-powered applications using OpenAI, Anthropic, Gemini, and other APIs.
- Implement resilient multi-API workflows with LiteLLM.
- Build RAG systems using embeddings, rerankers, query enrichment, and data indexing.
- Design and maintain data pipelines and vector storage using MongoDB Atlas Vector Search.
- Fine-tune LLMs on custom datasets for documentation and user-support applications.
- Collaborate with the YOLO development team on multimodal solutions.
- Automate deployment and testing processes with GitHub Actions.
Requirements
- At least 5 years of software engineering experience with a strong focus on AI and machine learning.
- Expert-level Python proficiency and deep learning experience, particularly with PyTorch.
- Experience building and deploying applications with OpenAI, Anthropic, Gemini, and DeepSeek APIs.
- Hands-on experience with complete RAG pipelines, vector embeddings, rerankers, and MongoDB-based indexing.
- Knowledge of LLM fine-tuning, prompt engineering, and performance optimization.
- Familiarity with MLOps and GitHub Actions-based CI/CD workflows.
- Strong interest in computer vision and understanding of object detection models such as Ultralytics YOLO26.
- Existing legal authorization to work in the UK.
- Strong problem-solving skills and ability to work in a fast-paced environment.
Benefits
- Full-time hybrid work in London with three days per week in the office and two days remote.
- 24 vacation days, a birthday day off, and local holidays.
- Flexible working hours.
- Equity packages.
- Apple MacBook Air or Pro, Apple Studio Display, and AirPods Pro 3.
- Opportunity to work on cutting-edge AI and open-source software.
Categories
About Ultralytics
Ultralytics builds computer-vision software centered on the Ultralytics YOLO family of real-time object detection models and supporting tools. Its business includes the Ultralytics HUB SaaS for dataset management, training, and deployment, along with open-source libraries, enterprise support, and commercial licensing. Founded in 2022 and headquartered in London, the privately held company raised a $30M Series A and focuses on helping developers and enterprises build and ship vision AI applications.