Lead AI Engineer (LLM & Agents)

Role overview A lead AI engineer is being recruited to own the backend and agent systems behind LLM-powered products for a major non-custodial cryptocurrency wallet serving a very large global user base. The role is end-to-end, covering agent workflow design, quality measurement, production reliability, and ongoing product direction for features that must work safely with on-chain and Web3 contexts.

Responsibilities - Design, build, and maintain multi-turn conversation systems, including context management, tool calling, retrieval-augmented generation, state management, and human handoff. - Define success criteria with product, design, and QA leads, and convert production failures into evaluation samples and regression tests, measuring impact before and after every release. - Diagnose root causes across the stack, including model judgment errors, knowledge gaps, abnormal tool responses, backend instability, and conflicting rules, applying the right fix rather than defaulting to prompt tweaks. - Own production reliability: timeouts, retries, graceful degradation, async jobs, and third-party dependency failures, ensuring every issue is traceable, every change is versioned, and every failure has a recovery path. - Shape product and technical decisions with reasoned positions on feasibility, cost, risk, and alternatives, proposing better approaches and breaking work into phased, shippable deliverables. - Communicate proactively on progress, blockers, ETAs, and next checkpoints, and follow through with post-launch tracking, documentation, and handover materials.

Requirements - Production LLM or agent experience at scale, having built and maintained features used by 10,000+ active users after launch. - Strong backend engineering fundamentals, with comfort owning production code, APIs, data flows, and incident troubleshooting. - Demonstrated evaluation discipline, including building evaluation sets, regression suites, and production monitoring, and the ability to separate real improvements from random fluctuation. - An ownership mindset that drives decisions, surfaces risks early, and considers a feature unfinished until it is verified in production. - Approximately one year of Web3 experience in a crypto, DeFi, or Web3 company such as a wallet, exchange, L1/L2 protocol, or DeFi platform.

Benefits and work setup - Fully remote, global role; candidates must have the right to work in their country of residence. - Flexible working hours within an international, flat organizational structure, alongside a globally distributed engineering team. - Competitive salary and benefits package, with reference to learning, career development, and challenging, fast-paced projects.