AI Builder

About the Role Curb is rapidly shifting to AI agents and agent-driven automation as a core operating capability. Across Finance, Marketing, Customer Service, and Product Operations, there are workflows that should generate insights, deliver outcomes, and provide early warning and monitoring — and the teams running them are ready to move. We’re hiring an AI Builder to lead the charge. This is a role at the intersection of business strategy, operations, product, and technology — and it must come from a product background. You will frame the problem like a product manager, understand the business like an operator, and design the solution like a product architect. Responsibilities Operate at the strategy / ops / product / tech intersection Frame the problem. Sit with Finance, Marketing, CS, and Product Ops leaders. Translate fuzzy business pain into a sharp problem statement, a hypothesis, and a measurable outcome. Prioritize like a product leader. Not every workflow deserves an agent. Pick the ones with real leverage — impact on revenue, cost, cycle time, or risk — and say no to the rest. Design the solution end-to-end. Data sources, skills, prompts, agent architecture, tool integrations, human-in-the-loop, failure modes, escalation paths, success metrics. The full product, not just the prompt. Build the straightforward ones yourself. Use Claude, Claude Code, Claude Agent SDK, and the broader agent tooling landscape to prototype, test, and iterate quickly. Partner with engineering to take the best ones to production. Pull in the team when the build gets heavy. You have access to engineering support for the harder builds — deep integrations, production hardening, performance work. You lead the spec, design, and outcome; the team helps you land it. Finance: variance analysis, revenue and spend modeling, close-process automation, anomaly detection, forecast agents that flag deviations before the CFO has to ask. Marketing: campaign performance monitoring, creative and content analysis, driver and rider sentiment tracking, attribution insights, content generation agents. Customer Service: ticket triage and routing, first-response and resolution agents, trend and escalation detection, quality monitoring across human and AI interactions. Product Operations: operational early-warning agents, ride and driver diagnostics, KPI watchdogs, executive summary generators, cross-system reconciliation. Deliver the three outcomes: Insights. Agents that read the data, draw the conclusion, and write the narrative — so leaders get answers, not charts. Outcomes. Agents that don't just automate work but move a metric. Every agent you ship traces to a P&L line, an SLA, a CSAT number, or a cycle time. Warning and monitoring. Agents that watch systems and data continuously, detect anomalies and drift, and escalate intelligently — turning reactive ops into proactive ops. Qualifications Product leader DNA. You have product manager experience — you’ve contributed to roadmaps, shipped features people use, and understand how to make the call on what not to build. You think in users, outcomes, and metrics — not features or deliverables. Engineering and AI builder fluency. You are comfortable in code and in the terminal. You understand APIs, data, and system trade-offs well enough to make good architectural calls. You design skills, compose tools, and engineer context. Hands-on with Claude, Claude Code, Claude Agent SDK or similar tools. You’ve built prototype agents and know enough to have a real conversation about how they work. Business and operational fluency. You are comfortable engaging with stakeholders across Finance, Marketing, CS, and Ops — understanding their workflows, asking the right questions, and translating business problems into product work. Bias to ship. You operate in ambiguity, pick up dropped balls, and ship in days. You move metrics, not tickets. Preferred Skills Agent and systems thinking. Exposure to designing multi‑step agentic workflows — understanding when a problem calls for a prompt, a tool‑using agent, a router, or a multi‑agent system. Eval mindset. Familiarity with evaluating agent quality — understanding the difference between deterministic and probabilistic evals and why measurement matters before shipping. MBA or equivalent business depth; prior strategy, consulting, or operator experience alongside your PM work. Experience with MCP (Model Context Protocol) and enterprise tool integration. Prior work in mobility, marketplaces, or multi‑sided platforms. Comfort with SQL and modern data stacks. Experience with RAG, long‑context agents, or structured output. Curb Mobility is an equal opportunity employer. #J-18808-Ljbffr Curb Mobility