AI Product Engineer - Full-Stack & LLMs

Amatra; Greater London; On-site; GBP 65,000 - 90,000; Full time

Job summary

Amatra is seeking a Product Engineer to design and build AI-powered features for our platform used by elite sporting organisations. You will work across the full stack—from data pipelines and LLM integrations to the interfaces that present insights to users.

Collaborating with founders and customers, you will translate needs into working software, own quality from prototype to production, and move fast with pragmatic decisions while iterating based on real usage.

Qualifications

• Strong engineering background preferred (but not required) in Computer Science, Mathematics, Software Engineering, Physics or Data Science.
• Proficient in Python, Java, C++, TypeScript/JavaScript; experience building web apps with React/Next.js.
• Familiar with AI/ML concepts and LLMs, embeddings; knowledge of classical ML is a bonus.
• Comfortable across the stack, from backend to frontend to infrastructure; bias toward shipping and iteration.

Responsibilities

• Build and ship AI-powered features across the full stack, from data layer to UI, including LLM-based workflows.
• Work closely with founders and customers to translate needs into working software.
• Own the quality and reliability of what you ship, from prototype to production.
• Move fast, make pragmatic technical decisions, and iterate based on real usage.

Skills

Python

Java

C++

TypeScript/JavaScript

Education

Computer Science

Mathematics

Software Engineering

Physics

Data Science

Tools

React

Next.js

Job description

Amatra is seeking a Product Engineer to design and build AI-powered features for our platform used by elite sporting organisations. You will work across the full stack—from data pipelines and LLM integrations to the interfaces that present insights to users.

Collaborating with founders and customers, you will translate needs into working software, own quality from prototype to production, and move fast with pragmatic decisions while iterating based on real usage.