AI Engineer (Backend)

Build the AI platform: Design, build and own the core platform that powers AI at Duffel. You’ll develop the orchestration engine, runtime and reusable primitives that allow intelligent agents to reason, plan and safely take actions across our products.

Define how AI is built at Duffel: Shape the technical direction for AI across the company. You’ll establish engineering patterns, development workflows and best practices that future AI systems and engineers will build upon.

Build trustworthy AI: Design retrieval, grounding and evaluation systems that ensure our agents are accurate, reliable and continuously improving. You’ll help us measure AI quality through robust evaluation frameworks rather than intuition alone.

Create long-term AI capabilities: Lay the foundations for memory and personalisation, building reusable customer context and optimisation loops that allow agents to become increasingly helpful while remaining predictable and safe.

Ship intelligent AI products: Design and build production AI agents that solve real customer problems, reasoning over complex travel workflows and safely interacting with Duffel’s APIs.

Raise the engineering bar: Work closely with engineers across the business, influencing architecture, mentoring others and helping establish Duffel’s long-term AI strategy., You’ll be building systems that operate against live travel bookings, payments and customer interactions where correctness, trust, latency and reliability matter. Your work won’t just power one assistant-it will define the AI foundations used across future Duffel products.

You’ll probably have experience across many (but not necessarily all) of the following:

• Significant experience designing and shipping production AI systems, agentic workflows or LLM-powered products.
• Experience building AI infrastructure such as orchestration frameworks, retrieval systems, evaluation pipelines or memory architectures.
• Strong software engineering fundamentals and the ability to design scalable, reliable systems regardless of language or framework. You’ll work across our stack as needed (Elixir on the backend and TypeScript on the frontend).
• Experience taking ambiguous technical problems from zero to one and influencing technical direction across teams.
• Strong judgement when balancing experimentation with long-term platform investment.
• Excellent communication skills and the ability to explain complex technical decisions to both technical and non-technical audiences.
• A collaborative mindset with a passion for solving difficult engineering challenges alongside others.

Bonus points if you have experience with:

• Travel technology, airline systems or online marketplaces.
• AI evaluation frameworks.
• Multi-agent systems.
• Production RAG architectures.
• AI observability and monitoring.
• Building developer platforms or internal AI tooling.

Travel is essential, whether it’s visiting family, starting a new adventure, or taking a career defining trip. Yet today, over 4 billion airline passengers rely on tech that hasn’t changed in decades.

At Duffel, we’re rebuilding the infrastructure underpinning the global travel industry. Backed by top-tier investors (Benchmark, Blossom, Index Ventures, and Kima Ventures) and born out of Y Combinator (S18), we’re simplifying complex systems to build the future of travel.

AI at Duffel

Travel is one of the most technically demanding domains for AI. Every booking involves multiple suppliers, APIs, payments, policies and real-world constraints. Building AI that can reason reliably, take safe actions and operate in production is far more than building a chatbot-it’s a systems engineering challenge.

We’re building the AI platform that will power intelligent travel experiences across Duffel. This includes the orchestration, retrieval, memory and evaluation systems that enable AI agents to reason over live bookings, interact safely with our APIs and continuously improve over time.

Our first application is a conversational travel assistant, but this role is about much more than a single product. You’ll help build the foundations that future AI products at Duffel will rely on, defining how we design, evaluate and scale AI across the company.

If you’re excited by solving complex engineering problems at the intersection of distributed systems, LLMs and production software, we’d love to hear from you.