Remote Dev Jobs
1 jobs available
Legalhero · 🇩🇪 Berlin, GermanyRemote57y ago
What You'll Do You’ll identify open AI opportunities, build them from start to finish, and bring them to real users. As a Senior AI Product Engineer, you’ll work across our three product squads: You’ll single-handedly turn unaddressed topics into products, hand them off once they’re up and running, and serve as a sparring partner for the squad PMs on all AI-related topics. For us, “senior” refers to your approach to work, not your years of experience. You’ll report directly to Product Leadership. Your Responsibilities Independent AI projects (about half of your time) You’ll identify high-impact AI opportunities in our legal workflows - in areas where no specific squad is responsible Validation happens quickly: discussions with attorneys, Legal Ops, and insurers; prototypes built in days You’ll bring the winning solutions into production yourself: routing, chaining, summarization, and agent-based loops You’ll hand off mature features to the squad - complete with documentation, evaluations, and a handover plan Support for the squads (the other half) Feasibility, costs, model selection, risks: You’ll answer these questions before every AI feature is launched You support the teams in designing evaluations, golden datasets, and guardrails You review concepts before they go into the roadmap - including prompt injection and compliance risks (GDPR, Section 203 of the German Criminal Code) You advance AI best practices: prompt engineering standards, LLM observability, and agent-based coding workflows What you bring Must-have Multiple production-ready AI/LLM systems that you’ve shipped to real users and taken responsibility for - no prototypes Demonstrable product ownership: from a vague problem to a finished product Strong Python skills and a solid foundation in software engineering (Git, testing, CI/CD, code review) In-depth experience with LangGraph, LangChain, or similar - including an understanding of what’s happening under the hood Proficiency with LLM provider APIs: fu