Fullstack Software Engineer – Automated Resolution

Описание

You would join the product team that builds the AI agents answering our 1M+ members' questions about their health insurance, aiming to increase the resolution rate and quality of answers.

Responsibilities

• Design and ship LLM-powered agents that understand a member's request, reason over their contract, claims and history, and take real actions on their behalf
• Build the tools, retrieval and context pipelines that give agents the right information at the right moment
• Make agents interoperate with other AI agents across Alan, such as Mo, by handing off conversations, sharing context and calling each other's capabilities
• Bring automated resolution to new channels, starting with AI phone support: real-time voice conversations
• Own evaluation: build datasets, automated evals and quality monitoring to ensure an agent is right before members see it
• Measure and improve resolution rate, answer quality and member satisfaction in production, iterating fast on prompts, models and architecture
• Work hand in hand with care experts and product to identify the next topics the AI can fully resolve

Requirements

• 5+ years building scalable software systems
• Strong backend experience (Python is a plus)
• Hands-on experience building products with LLMs (agents, tool calling, RAG, prompt engineering, evaluation) - ideally shipped to production
• An evaluation mindset: you know how to measure whether an AI system is actually good, and how to make it better
• You love turning complex problems into elegant solutions
• You care about the "why" behind what you build, not just the "how"
• You stay on top of a fast-moving AI landscape and are pragmatic about what really works
• Humble and still willing to grow
• Healthcare technology excites you
• You understand that insurance isn't just about money - it's about access to care

We offer

• A stimulating environment and perks ensuring Alaners are happy, efficient and spend only high-quality time with co-workers
• Innovative working methods and a strong culture guided by cultural values
• The stack is React, Python/Flask, PostgreSQL in a monorepo
• The possibility to join a different engineering team based on where we think you'll have the most impact

Hiring process

• The process is very simple: those who make product decisions are the same ones who build them