ironquill.tech/board

$ cat jobs/customer-operations-ai-engineer-tread-3e89b04d68ce.json

Customer Operations AI Engineer

Tread·US·San Francisco·mid
Apply on ashby → Get AI match score →
COMPANY Tread is an AI-native vertical SaaS platform transforming construction materials logistics. The company crossed $1Bn in monthly delivered load value in March 2026. The platform serves Haulers, Producers, and Contractors, optimizing truck routing and delivery while providing customer service through AI agents. Tread is Series A-funded by Mucker Capital. Why This Role Exists Tread's customers operate complex, high-stakes logistics, and you can't onboard, support, or grow them from a desk. This is the customer-side counterpart to our product-side Forward Deployed Engineer: instead of shipping product code, you build and own the AI agents that run support, onboarding, and success, deployed alongside real customers with feedback loops measured in hours. As AI takes on the volume of customer work, this role lets us scale customer operations without scaling headcount. It sits in Customer Operations, reports to the VP of Customer Ops, and is measured on customer outcomes. What You Own - Agentic Support: Build the agents that resolve support end-to-end. Anything AI can answer gets answered by AI, with a human audit. First targets are our two highest-volume drivers: login issues and multi-driver phone-number issues. Own the AI support layer (Fin+ the Mintlify knowledge base behind it): decide what's safe to automate, validate against real prior conversations before go-live, and keep the content trustworthy. - Agentic Onboarding: Build end-to-end agentic onboarding flow, customized per customer. Vendor onboarding comes first, then product onboarding and case reporting. The agent tracks each customer's progress, delivers the right step at the right moment, and escalates only the stalls. - Agentic Success & Enablement: Map every post-onboarding job to be done (QBRs, monthly check-ins, renewal prep) and automate it, including voice-based check-ins. Build account-health and usage monitoring that flags risk before a human would see it, plus self-serve enablement that reduce