Senior AI Software Engineer

We are looking for a Senior AI Software Engineer to build and improve the AI systems already used across Kiwi's products and operations. Our work includes LLM powered chat and voice experiences, agents that use tools to complete tasks, and automations connected to real customer and business workflows.

You will work across AI and backend engineering: designing agent behavior, integrating LLMs with our services and data, building reliable APIs, and improving how these systems perform in production. This is a hands-on role for someone who has shipped LLM applications beyond prototypes and takes responsibility for their accuracy, latency, cost, and reliability.

You will collaborate with Product, Design, Operations, and Engineering to understand the problem behind each use case, decide where AI adds value, and turn that decision into software that works for real users.

• →

  Build and operate LLM powered features, including conversational agents, voice experiences, and workflow automations.

• →

  Design how agents use tools, retrieve information, maintain context, and hand off to people or other systems when needed.

• →

  Integrate AI capabilities with Kiwi's backend services, APIs, product flows, and internal tools.

• →

  Improve the quality of AI behavior through evaluations, production feedback, testing, and iteration.

• →

  Investigate and resolve issues involving incorrect responses, failed tool calls, latency, cost, or unexpected behavior in production.

• →

  Design clear safeguards for actions that affect customers or business processes, including validation, permissions, and human confirmation where appropriate.

• →

  Contribute to architecture and code reviews, share practical engineering standards, and help other engineers build dependable AI features.

• Strong software engineering experience, with the ability to design, build, and operate backend services and APIs in production.

• Hands-on experience building and shipping applications with LLMs, including tool calling, agent workflows, and integration with external systems.

• Experience improving LLM applications using evaluations, traces, user feedback, or production metrics. You can explain what failed, how you measured it, and what you changed.

• Strong TypeScript and Node.js experience. Working knowledge of Python and the ability to contribute to Python services when needed.

• Experience designing for reliability in asynchronous workflows, including failure handling, retries, observability, and recovery.

• Ability to work with ambiguous product problems, challenge assumptions, and propose practical solutions rather than simply implement a specification.

• Sound judgment about when an AI system can act autonomously and when it needs validation or human review.

• Clear communication and ownership across Engineering, Product, and business teams.

• Experience with voice agents, speech to text, text to speech, or real-time conversational systems.

• Experience with retrieval augmented generation, knowledge bases, or document processing.

• Experience with AI workflows in fintech, lending, payments, customer support, or other environments with sensitive customer interactions.

• Experience improving the cost and performance of LLM applications at scale.