AI Engineer (Process automation via AI)

An internal platform for the company. It is already live, people use it every day, and it has an AI assistant inside. A second application on the same core is starting now. The team is small: whoever writes a feature ships it and looks after it afterwards.Nobody hands you a finished spec. A task arrives as someone's problem. You go to that person, find out what is really needed, and write the description of how it should work yourself. The code, the documentation and the assistant all follow from that description. Then you decide how to build it, and the result is yours.The AI work here is engineering, not research: agents, tool calling, search over documents, checking quality. All of it inside a product that holds employee data, so mistakes are expensive.Stack: TypeScript everywhere, pnpm + Turborepo monorepo; Nuxt / Vue, Pinia, i18n, PWA; Nitro / h3, Prisma + PostgreSQL, better-auth, Zod; Temporal; Mastra, OpenRouter and AG-UI, search on pgvector; a desktop app on Tauri + Rust; Vitest, Playwright, Testcontainers; Kubernetes, Helm, Terraform, GitLab CI.Requirements:You can work from a problem instead of a spec. You ask questions first, then write down what you propose to do. You can say no when a task is being solved the wrong way.4+ years in web development, both front end and back end. You have built a product that reached real users.Good TypeScript, Vue (or React), and a Node backend.You know databases: design a schema, write a query, run a migration on live data.You have worked with LLMs in production: tool calling, structured output, search over documents. You can check quality — change a prompt and show in numbers that it got better.You think about access rights and audit without being reminded: the system holds employee data, and a model can propose more than it should.You write tests as you go, not afterwards.Responsibilities:Work out what a request really is: talk to the person who brought it, understand what is needed, build that;Build features whole: data schema, interface, tools for the assistant, permissions, migrations, tests, deploy;Grow the assistant: many tools, finding the right one on the first try, agent scenarios, safety limits, quality checks;Keep things running: integrations, CI, Helm, Terraform, the cluster, Postgres;Take tasks from different teams across the company, not from one department.We offer:High salary (plus performance bonuses and salary revision);Remote work;Work schedule: Mon-Fri (8 hours per day) with flexible starting hours;24 days holiday leave;Corporate English lessons;Exciting work challenges that allow you to grow to your full potential;A strong team of like-minded professionals who will be by your side to accomplish ambitious projects, stimulate your professional development and bring experience.