AI Software Engineer

CloudMargin; Greater London; On-site; GBP 110,000 - 160,000; Full time

Job summary

CloudMargin is seeking a senior software engineer with deep AI interest to shape our AI-native product strategy. You will lead AI coding agents, orchestrate the engineering pipeline and deliver secure, production-ready code in a small focused team.

Ideal candidates have 10+ years in software development, strong NodeJS/TypeScript or Python, React experience, and hands-on AI experience with agents or LLMs in regulated or financial contexts.

Qualifications

• 10+ years in software development as a baseline.
• Proven experience with NodeJS (TypeScript) or Python and front-end React.
• Hands-on with AI coding agents or AI-powered features.

Responsibilities

• Build AI and agent capabilities into products and services.
• Deliver end-to-end engineering outcomes with robust testing (TDD).
• Orchestrate AI coding agents within a human-owned pipeline.
• Collaborate with engineers and product teams to raise quality.

Skills

NodeJS

TypeScript

Python

React

AI agents

LLMs

AWS

DevOps

TDD

REST/SPA

Education

Computer Science degree

Tools

AWS

CI/CD

Docker

Kubernetes

Job description

Primary Purpose of Role

You will combine deep software engineering experience, excellent judgment and a demonstrable interest in AI to help shape the way we build our next generation of products. Initially working in a small, focused team, you will help lead and progress the development of our foundational AI-native / AI-first engineering approach, embedding and amplifying the power of AI to build and deliver sophisticated planned and future products.

Core Accountabilities

You will build AI and agent capabilities into our products and services, and exercise the judgment that keeps consequential decisions in human hands. You will deliver engineering outcomes end-to-end, keeping abreast of industry advancements and leveraging proven practices to turn designs into secure, high-quality production-ready code.

Experience

• 10+ years or equivalent experience in software development.
• Proven experience in either NodeJS (TypeScript) or Python
• At least 2+ years experience with React or a similar frontend framework.
• Hands-on experience shipping production code with AI coding agents or assistants (e.g. Claude Code, Cursor or equivalent).
• Experience integrating LLMs or building agentic / AI features into a product (desirable).
• Experience in a regulated or financial-services environment (desirable).

Role Scope

You will operate as an orchestrator of AI coding agents within a human-owned engineering pipeline, accountable for everything that reaches production.

• The Engineering Harness: Initially working with a small-project team, build and maintain the shared conventions and guardrails that make agents safe and useful. Keep the team’s single source of truth intact and continuously design out recurring failures.
• Agent Orchestration & Delivery: Ship production software by directing AI coding agents, owning the merged result end-to-end.
• Quality, Review & Verification: Ensure agent-generated code meets acceptable standards and architectural decisions by applying strong testing discipline including TDD.
• AI & Product Capabilities: engineer AI and agent features into the product.
• A Human Agent for Change: working collaboratively with engineers and product teams to solve problems and align on best practice, documenting decisions throughout. Influencing our next generation of engineers and mentoring on AI practices, experiences and outcomes to raise the bar on quality and outcomes.

Skills

• Orchestrating AI coding agents: task decomposition, context and prompt engineering, defining acceptance criteria and stop conditions.
• Rigorous review and verification of AI-generated code, backed by strong testing discipline (TDD, plus performance and security testing). Knowing when to keep a human in the loop.
• Advanced modern JavaScript / TypeScript, NodeJS and React.
• Web development principles (REST, SPA…).
• Microservice architecture, AWS and DevOps.
• Building LLM-powered features: tool-calling, retrieval-augmented generation (RAG), evaluation / eval harnesses and observability (desirable).
• Familiarity with model-agnostic tooling and routing, e.g. a proxy / gateway layer (desirable).
• Agile manifesto and SCRUM.
• Advanced written and verbal communication.
• Passion for delivering high-quality software and a positive approach to continuous improvement.

Education / Professional Qualifications

Computer Science or Science-based degree