AI Product Engineer

Understanding Recruitment; Royal Borough of Greenwich; Hybrid; GBP 70,000 - 110,000; Full time

Job summary

A London-based startup building AI-powered design tools is seeking an AI Product Engineer to bridge users and our AI systems. You will contribute across frontend, backend and applied AI to turn ideas into working, structured workflows and help engineers inspect and refine results.

You will translate natural language requests, images and drawings into precise specifications, design agentic systems, evaluate models, and create end-to-end features from design through implementation in a small,

Qualifications

• Strong product-engineering fundamentals.
• Deep understanding of LLM behaviour beyond prompting.
• Experience building agents that use tools and backend services.
• Sound judgement on model selection, evaluation and reliability.
• Ability to translate ambiguous user needs into product features.
• Capacity to work across frontend, backend and applied ML in a small team.

Responsibilities

• Build polished applications that help engineers describe, generate, and refine designs.
• Translate natural-language requests, images, and drawings into structured engineering specifications.
• Design agentic systems that interact reliably with tools, backend services, and CAD geometry systems.
• Evaluate models, investigate failures, and decide when adaptation or fine-tuning is needed.
• Create workflows that let users review, correct, and guide AI-generated outputs.
• Own features across frontend, backend, and applied AI, from initial design through implementation.

Skills

Product engineering

LLM behaviour

Agents with tools

Model selection

Frontend/Backend ML

User story translation

Job description

AI Product Engineer - AI/Hardware/Manufacturing

Location: Central London, primarily office-based with some flexibility.

We're partnering with an AI startup building software to help engineers design and develop physical products faster. Combining machine learning with complex engineering software, their platform helps bring newhardware ideas closer to production.

They're looking for anAI Product Engineer to build the interface between users and their AI-powered design systems. This is a hands-on opportunity to work across frontend, backend, and applied AI, helping engineers turn ideas into structured workflows and inspect and refine the results.

What you'll be doing

• Build polished applications that help engineers describe, generate, and refine designs.
• Translate natural-language requests, images, and drawings into structured engineering specifications.
• Designagentic systems that interact reliably with tools, backend services, and CAD and geometry systems.
• Evaluate models, investigate failures, and decide when adaptation or fine-tuning is needed.
• Create workflows that let users review, correct, and guide AI-generated outputs.
• Own features across frontend, backend, and applied AI, from initial design through implementation.

What we're looking for

• Strong product-engineering fundamentals and experience delivering user-facing software.
• A deep understanding ofLLM behaviour, beyond prompting and basic API integration.
• Experience building agents that use tools and interact with backend services.
• Sound judgement around model selection, evaluation, and reliability.
• Confidence translating ambiguous user needs into practical product features.
• The ability to work across frontend, backend, and applied ML within a small, highly technical team.

Experience that would stand out

• React, TypeScript, Python, and modern application architecture.
• Structured generation, tool use,RAG, model adaptation, or fine-tuning.
• Multimodal systems involving text, images, drawings, or technical inputs.
• Evaluation frameworks, observability, and testing for AI systems.
• CAD, geometry, engineering software, or human-in-the-loop workflows.

We're interested in engineers who combine strong coding skills, a practical understanding of AI, and thoughtful product decisions to build software that works reliably for users.