We are seeking a Lead AI-Native Engineer to raise a team's or customer's AI maturity and prove it with numbers. This role designs and ships AI systems, with success measured by what changed for others: engineers who now maintain their own skills, a customer team that moved from pilot to production, and adoption and quality metrics that improved because of how the work was introduced. Responsibilities Lead AI adoption for a squad or customer team: choose the first wins, handle resistance (vendor teams, sceptical senior engineers, delivery pressure), and build trust before changing process Design the delivery pipeline and its human gates, choose the model mix, MCP servers, tool sets and skill registry, and set the tool-restriction policy Track adoption and quality metrics (acceptance rate, PR cycle time, defect density, AI-touched PR ratio), tune the workflow from them, and report outcomes to leadership Set how the team or product tracks and controls AI cost: cost per task or outcome, budgets and alerts Coach engineers up a level and review the skills and agents they create Represent the AI delivery model to client and business stakeholders: explain trade-offs and hold the delivery model under scope pressure Document the operating model so a peer can run it without you Harvest reusable skills, patterns and reference architectures back to the wider organization Keep a documented path towards more autonomous delivery, and decide which steps are safe to remove human review from Keep designing and shipping AI systems: multi-stage LLM pipelines and agents with a curated tool set, backed by evaluation Requirements 8+ years of professional software engineering experience 3+ years of hands-on experience with LLM or agent systems 2+ years in a lead, architect, mentoring or customer-adoption capacity Showcase of moving 3 or more engineers up a level, or moving a customer team from pilot to production use, with numbers to show it Background in a production AI system with quantitative outcomes (accuracy, cycle time, cost, hours saved, adoption), the failure modes encountered, and the structural fixes Expertise in AI work that runs without daily attention, multi-step LLM or agent design with tool restriction and schemas, an evaluation approach, and structural fixes for wrong output Understanding of trade-offs named: what was chosen, what was rejected and why, ideally across more than one project Track record in adoption: first wins, metrics, tuning, and how resistance was handled Proficiency in a cost-control approach that others follow, including a measured reduction that kept quality Familiarity with a documented operating model that a peer can run without the candidate Skills in clear communication with outcomes and lessons from real situations, including with skeptical audiences Nice to have Background in reusable patterns adopted by other squads or engagements, and multi-agent topologies designed and run in production, with observability (token spend, hallucination rate, false-positive ratio, retry counts) Experience in governance and compliance for AI (guardrail policies, approval flows, audit trails) Familiarity with pre-sales or client-facing work Understanding of autonomous delivery pipelines, with concrete next steps Showcase of conference talks, internal courses or written guidance on AI adoption
