Senior AI Infra Engineer - Model Training & Inference

intercom; Greater London; Hybrid; GBP 90,000 - 150,000; Full time

Job summary

Fin, now part of Salesforce, is building AI infrastructure to train and serve next‑gen models at scale in London and beyond. You’ll join a small, highly technical team focused on training pipelines, low‑latency inference services, and GPU optimization across the stack.

We’re seeking Senior+ AI Infrastructure Engineers with 5+ years of software engineering, experience in model training or inference, and strong fundamentals.

Qualifications

• 5+ years of software engineering experience with strong delivery track record.
• Degree in Computer Science, Computer Engineering, or equivalent.
• Hands-on in one or more: model training, model inference at scale, or low-level GPU work.

Responsibilities

• Implement and scale training pipelines for large transformer/LLM models.
• Build and optimize inference services with low latency and high reliability.
• Tune GPU kernels and optimize training & inference performance.
• Collaborate with ML scientists to productionize new methods.
• Mentor and develop other engineers; raise technical standards.

Skills

Model training

Model inference

Low‑level GPU work

CUDA

Triton

Kubernetes

Python

Education

Degree in Computer Science/Engineering

Tools

CUDA

Triton

Kubernetes

AWS

Job description

Fin, now part of Salesforce, is building AI infrastructure to train and serve next‑gen models at scale in London and beyond. You’ll join a small, highly technical team focused on training pipelines, low‑latency inference services, and GPU optimization across the stack.

We’re seeking Senior+ AI Infrastructure Engineers with 5+ years of software engineering, experience in model training or inference, and strong fundamentals.