高级全栈工程师，AI构建者
Senior Full Stack Engineer, AI Builder
保险是世界上最大的行业之一，也是最令人痛苦的交互体验之一。纸质表格、缓慢的客服中心、三周的核保周期、消失在队列中的索赔。我们正在基于现代平台和真正能完成工作的AI代理重新构建这一切，而不是假装能做。
你不会处理工单。你会决定要构建什么，在周一进行原型设计，周五将草稿发送给真实客户，并根据真实使用情况进行改进。我们期望你推动产品、代码库以及我们的工作方式。
为什么这个职位存在
我们寻找一位将AI视为一种创造材料，而非路线图上的一个功能的人。保险的下一年将由现在定义。哪些内容会被自动化，哪些会被重新构想，哪些会被抛弃。我们想要那些希望参与其中并推出解决方案的建造者。
团队很小，使命很大。如果你以往最高能量的工作来自于从头到尾解决真实问题，这就是你的位置。
你将做的事情
- 负责结果，而不是工单。每个冲刺周期选择对客户最重要的杠杆，并将其推向上线。自己发现下一个杠杆。
- 以AI为主要工具，而不是作为聊天机器人的附加功能。OCR、文档理解、生成、代理循环、评估工具。你决定每个部分的位置。
- 快速发布原型，然后使其具备生产级质量。第三天的可用演示胜过第十天的完美计划。
- 直接与客户合作。你将看到真实的索赔、真实的核保界面、真实的经纪人工作流程。你将观察用户遇到的死胡同，并决定要修复什么。
- 提升代码库的标准。审查、重构、可测试性、可观测性。不是作为政策，而是作为习惯。
- 推动团队前进。技能、代理、内部工具、工程标准。提出它们，构建它们，让它们被采用。
你是什么样的人
- 你注重交付。没有执行的想法对你来说毫无意义。
- 高度自主。你不会等待工单被分配。你看到值得做的事情，就会开始行动。
- 有坚定的观点，但保持开放。你在信息不完整的情况下做出决策，并在证据变化时改变主意。
- 对领域充满好奇。保险很复杂，充满奇怪的激励机制。这让你着迷，而不是困扰你。
- 精通AI。你曾用LLMs或代理实际交付过项目。你知道魔法的尽头在哪里，以及在哪里需要仔细工程。
- 关心另一端的人。正确的保险能给人带来安心。
查看英文原文
Insurance is one of the world's largest industries and one of the most painful to interact with. Paper forms. Sluggish call centres. Three-week underwriting cycles. Claims that vanish into queues. We are rebuilding all of it on top of a modern platform and AI agents that actually do the work, not pretend to.
You will not be picking up tickets. You will be deciding what to build, prototyping it on Monday, shipping a draft to a real customer by Friday, and improving it from real usage. We expect you to push the product, the codebase, and the way we work.
Why This Role Exists
We are looking for someone who treats AI as a creative material, not a feature on a roadmap. The next year of insurance is being defined right now. What gets automated, what gets reimagined, what gets thrown out. We want builders who want to be in that room and ship the answer.
The team is small. The mandate is large. If your highest-energy work has historically come from solving real problems end to end, this is the seat.
What You Will Do
- Own outcomes, not tickets. Pick the biggest lever for our customers each sprint and drive it to live. Surface the next one yourself.
- Build with AI as a primary tool, not a chatbot bolt-on. OCR, document understanding, generation, agent loops, evaluation harnesses. You decide where each fits.
- Ship prototypes fast, then make them production-grade. A working demo on day three beats a perfect plan on day ten.
- Work directly with customers. You will see real claims, real underwriting screens, real broker workflows. You will watch users hit dead ends and decide what to fix.
- Raise the bar on the codebase. Reviews, refactors, testability, observability. Not as policy, as habit.
- Push the team forward. Skills, agents, internal tools, engineering standards. Propose them, build them, get them adopted.
Who You Are
- You ship. Ideas without execution do not count for you.
- High agency. You do not wait for a card to be assigned. You see what is worth doing and you start.
- Strong opinions, loosely held. You make calls with incomplete information and change your mind when the evidence shifts.
- Curious about the domain. Insurance is messy and full of weird incentives. That intrigues you, not annoys you.
- Fluent with AI. You have shipped something real with LLMs or agents. You know where the magic ends and where you have to engineer carefully.
- Care about humans on the other end. Insurance done right gives people peace of mind in their worst weeks. That is the point.
Your Toolbox
We use Node.js + TypeScript + Nest.js on the back end, Vue (and some React) on the front end, GraphQL, MongoDB, Docker, AWS. AI work spans OpenAI, Anthropic, Google, plus a growing set of in-house skills and agents. You do not need to have used every piece. You do need to be the kind of engineer who picks up a new tool and is shipping with it within the week.
What We Expect at This Level
- 6+ years building production systems. Track record of complex, customer-facing software shipped, not just maintained.
- Real AI experience. You have built and shipped at least one substantive AI-driven feature. POCs that died in a notebook do not count here.
- Strong fundamentals. OO, design patterns, testing, web security, performance. You do not need a refresher.
- Cloud comfort. AWS, Azure, or GCP. At least one in depth, the rest enough to hold a conversation.
- Communication. You can explain a tradeoff to a non-technical stakeholder in two minutes and to an engineer in one.
Not Suitable for the Role If…
- You prefer tickets assigned to you over choosing what to build.
- You measure your work in lines of code, not outcomes.
- You see AI as a buzzword to put on your CV rather than a tool you have actually used in anger.
- You want a process-heavy environment with detailed specs handed down from product.
- You are looking for a maintenance role on a finished platform.
Why CoverGo
Insurance is the rare industry where AI can save people hours of their worst weeks. We are already shipping AI-powered claims, underwriting, distribution, and document automation to enterprise carriers across five continents. The platform is real. The customers are real. The next chapter is yours to help write.
Why You'll Love Working Here
- 100% Remote Work
- 15 Days of Annual Leave
- Annual Performance Bonus
- Remote Work Allowance
- Anniversary Bonus
- Health Insurance
- Company Activities and Events
- Learning and Development Plan
CoverGo Company Video
By submitting your application, you confirm that you have read, understood, and accepted the content of CoverGo’s Privacy Notice and you consent to the processing of your data as part of this application.
Originally posted on Himalayas