About the Role
Clutch builds AI agents that work on behalf of credit unions — starting with Emma, our AI collections agent. In a recent deployment, Emma handled 60,000+ conversations with members (the credit union term for customers) in her first twelve weeks on the job — roughly 24 collector-years of work, delivered with a consistency no human team can match. And we’re just getting started: Emma is the first of a growing family of member-facing AI agents.
As AI Operations Manager, you will own the full post-sale journey for these agents — from the moment a credit union signs to the moment the agent is performing at scale, and every quarter after. That means designing each agent's behavior with the credit union (conversation design, escalation rules, negotiation boundaries, tone), running live testing, driving go-live, and then owning ongoing performance: monitoring results, catching drift, and proving outcomes in executive business reviews.
You are the expert the credit union is paying for. Most credit unions we work with have never deployed an AI agent — they don't know what good looks like. Your job is to walk in with a strong, opinionated recommendation for how the agent should behave, grounded in what has worked across deployments, and let the credit union adjust from there. And because this function is new, your second job is to build it: the playbooks, the defaults, the monitoring infrastructure, and eventually the team.
About the Team
You'll join the Special Projects team, working daily with the small, fast-moving pod that builds and ships our AI agents — product, engineering, and dedicated AI sales specialists. This is the part of Clutch operating closest to the frontier: the agents you'll run are being co-developed on-site with some of the largest credit unions in the country, and the roadmap changes based on what you learn in the field.
The team's operating style is direct and evidence-driven. Field intelligence you gather flows straight back into how we build, price, and sell — you will have a real voice in product decisions, not a feedback form.
You'll start as the first dedicated post-sale owner for our AI agents. As the agent portfolio and customer base grow, you'll define the roles, playbooks, and hiring plan for the team that scales this function — and lead it.
What You’ll Do
Within 1 months, you will:
Become an expert in the business you're stepping into: how credit unions operate, how a collections department runs day to day, the compliance rules that govern member contact, and what a past-due member's experience actually looks like
Learn Emma end-to-end: the system prompt architecture, the full configuration surface (verification protocols, escalation triggers, constraint language, payment negotiation boundaries, tone), and the implementation process from kickoff to go-live
Shadow active implementations and live customer deployments, including dry-run testing sessions with credit union collections teams
Map the current post-sale journey and identify where context is lost, where defaults are missing, and where the process breaks
Within 3 months, you will:
Own Emma implementations end-to-end as the customer's single point of contact: conversation design, business rule definition, live testing facilitation, and go-live sign-off
Own post…
