Senior LLM Engineer — Hybrid AI for Vision & NLP

Ultralytics; Greater London; Hybrid; GBP 75,000 - 110,000; Full time

Benefits offered by this job

Equity

Hybrid schedule

Birthday off

MacBook Pro/Air

Studio Display

AirPods Pro 3

Job summary

Ultralytics in London is seeking an experienced LLM Engineer to lead the integration of advanced language-model capabilities across our AI platform. You will collaborate with the YOLO development team on multimodal projects and help design robust, scalable ML solutions.

This full-time role offers a hybrid work model (3 days in-office, 2 remote). You will work with OpenAI, Anthropic, Gemini, and other APIs, build RAG systems, fine-tune models, and deploy them via CI/CD pipelines using GitHub

Qualifications

• 5+ years of software engineering with a focus on AI/ML.
• Expert-level Python and PyTorch.
• Experience with LLM APIs (OpenAI, Anthropic, Gemini).
• Experience with full RAG pipelines and vector databases (MongoDB).

Responsibilities

• Develop and scale LLM-powered apps using multiple APIs.
• Implement multi-API workflows with LiteLLM.
• Build Retrieval-Augmented Generation systems with embeddings and rerankers.
• Design data pipelines and MongoDB Atlas Vector Search.
• Fine-tune LLMs on custom datasets for docs and user support.
• Collaborate with YOLO development team on multimodal solutions.
• Automate deployment and tests with GitHub Actions CI/CD.

Skills

Python

PyTorch

LLM APIs

RAG pipeline

MLOps

MongoDB

Fine-tuning

Prompt engineering

CI/CD

Computer vision

Tools

GitHub Actions

LiteLLM

Voyage AI

OpenAI API

Anthropic API

Gemini API

MongoDB Atlas Vector Search

PyTorch

Job description

Ultralytics in London is seeking an experienced LLM Engineer to lead the integration of advanced language-model capabilities across our AI platform. You will collaborate with the YOLO development team on multimodal projects and help design robust, scalable ML solutions.

This full-time role offers a hybrid work model (3 days in-office, 2 remote). You will work with OpenAI, Anthropic, Gemini, and other APIs, build RAG systems, fine-tune models, and deploy them via CI/CD pipelines using GitHub