motorway - Principal Generative AI Engineer

• Set the overarching architecture for GenAI at Motorway, including reusable patterns for prompt design, RAG (hybrid retrieval, chunking, semantic caching), agent orchestration, evaluation frameworks (LLM-as-judge, golden datasets, adversarial prompts), guardrails and observability that other teams build on • Design and build the hardest single pieces of work directly, taking on the problems nobody else is well-placed to solve and leading our highest-risk, highest-reward technical bets across the GenAI surface area • Hold sign-off on architectural decisions for materially complex GenAI systems, with the credibility to veto implementations that don't meet our agreed technical, safety or quality bar • Advise the Director of Data and AI on build-vs-buy calls for AI infrastructure (vector stores, eval platforms, AI gateways, model routers, agent frameworks), and own the technical rationale behind them • Drive cost, latency and throughput optimisation across our production GenAI systems, with the trade-offs documented and explicit • Define our principles for emerging paradigms, including multi-agent protocols, multimodal systems and agent-to-agent communication • Define our responsible AI principles and keep us aligned with the EU AI Act, NIST AI RMF and ICO guidance ahead of regulatory deadlines • Own opinionated relationships with our model and infrastructure providers, with the credibility to influence vendor roadmaps where it matters to us • Coach engineers across the function, shape how we hire and review, and set the cultural bar for safety, evaluation and production quality • Identify what Motorway should stop pursuing with the same clarity as what we should start, and represent Motorway externally through publications, conference talks and meaningful open-source contributions, used as a multiplier for hiring and partnerships