AI Software Engineer (Senior or Staff)

Ledgy; Zürich; Hybrid; GBP 90,000 - 130,000; Full time

Job summary

Ledgy is seeking a Software Engineer to join our AI Systems team, building the Ledgy Agent, MCP server, and related AI foundations. You will collaborate with engineers, product managers, and designers to deliver correct, scalable AI features that live in our equity management platform.

You will work across the Ledgy web app and agent-heavy workflows, ensuring reliability, observability, and safe embedding of AI in customer workflows.

Qualifications

• 5+ years of experience as a Software Engineer with strong TypeScript/Node fundamentals across backend and frontend.
• Built and shipped agentic or LLM-based software to production, such as an agent, an assistant, or a chat product.
• Experience debugging failures in real conversations and optimizing cost and latency as usage evolves.

Responsibilities

• Design and build Ledgy Agent, the tools, skills, context, and orchestration behind it.
• Shape Ledgy MCP server integration so capabilities work where needed and ensure safety.
• Measure quality of the agent through evaluations, tracing, and real user feedback.
• Design scalable foundations for AI features across the Ledgy web app and future surfaces.

Skills

TypeScript/Node

LLM-based projects

Agent framework

Frontend/Backend

Job description

At Ledgy, we’re on a mission to make Europe a powerhouse of entrepreneurship by building a modern, tech-driven equity management and financial reporting platform for private and public companies. In 2026, we aim to be the leading provider for European IPOs and reporting for share-based payments. We are a value-based company with a core focus on being humble, transparent, ambitious and impactful, all in order to delivery the best experience for our customers and end users. We are proud to partner with some of the world’s leading investors. New Enterprise Associates led our $22m Series B round in 2022, with Philip Chopin joining Sequoia’s Luciana Lixandru on our board. We were founded in Switzerland in 2017 and today we operate globally from offices in Zurich and London. We encourage diversity and are an international team coming from 26 different countries and speaking 25 different languages.

About the Role

As a Software Engineer on our AI Systems team, you'll build the AI foundations and product experiences of Ledgy: the Ledgy Agent, the tools and skills it runs on, our MCP server, and the platform capabilities that let other teams embed AI safely in their domains. You'll collaborate with fellow engineers, product managers, and designers in a new and fast-moving product area where correctness matters as much as speed - our customers need equity data they can trust.

Location:

We hire remotely across the UK, Spain, Germany, Poland, and Romania. For candidates based in London, we use a hybrid model with 2 days in the office per week.

The team you'll join:

AI Systems - Develops Ledgy's AI foundations and product experiences, from AI agents and AI-assisted workflows to the platform capabilities that help other teams embed AI safely in their domains. The team works in a new and evolving product area, balancing rapid iteration with reliable execution, evaluation tooling, and responsible automation.

What You'll Work On

Help build the future of how companies manage equity. Ledgy Agent. You'll design and build the system behind our in-app agent, so equity administrators can do through AI what they do manually in the app today. Ledgy Agent takes real actions and runs real analysis on live equity data, like onboarding a stakeholder end to end, and administrators depend on its accuracy and observability. You'll design the tools, skills, context, and orchestration behind it, and how it decides to handle each request.

Ledgy MCP. Not everyone wants to open Ledgy to get things done. Our MCP server lets customers bring equity into the AI agents and tools they already use. You'll shape it alongside the in-app agent, so a capability built once works in both places, and you'll decide what we expose and how we keep it safe.

Measuring and improving the agent. You'll design and build the system we use to measure the agent's quality and learn from how customers use it: evaluations, tracing, and feedback from real conversations. You decide what "good" means for customers, how we measure it, and how what we learn changes what we build next.

Scaling. The Ledgy web application, the agent, MCP, and surfaces we haven't built yet all need the same capabilities. You'll design the foundations that let us build AI features once and reuse them everywhere, and that let other teams add AI to their domains without compromising correctness. You'll also design how it runs as usage grows: the backend infrastructure, cost, and latency behind the agent and MCP, and how we keep them reliable and observable. You'll own real parts of this and have a say in where we invest. You'll also spend time with customers to see how administrators actually use the agent, and use that to decide what to build next.

You’ll Thrive in This Role

If You Don't worry if you don't tick every box. These are the things that would make you an especially good fit.

• You have 5+ years of experience working as a Software Engineer, with strong TypeScript/Node fundamentals across back end and front end
• You've built and shipped agentic or LLM-based software to production, such as an agent, an assistant, or a chat product, and you've been the one running it.
• You've debugged failures in real conversations and kept cost and latency in check as usage and models changed.
• You've designed the agent itself: tools and skills, context and prompt engineering, state, streaming, and the agent loop.
• You know at least one agent framework or runtime well and can say where it works and where it doesn't.
• You improve these systems with evidence. Evals, regression sets, tracing, and feedback from real usage tell you whether a change helped, and you use them to decide what to do next.
• You keep up with what the foundation labs and the wider community are learning about building agents.
• You try new techniques early and judge which ones make our product better and are worth adopting now, and which should wait.
• You take part in deciding what we build. You read the data, talk to customers, and push for what you…