Want to know if this job is worth applying to?
San Francisco, California, United States
Onsite
At a glance: San Francisco, on site, with regular time in Atlanta · $220K to $260K, plus 15 to 20% bonus and meaningful equity · Runs customer operations across San Francisco, Atlanta and our team in Pakistan.
Harper is an AI-native commercial insurance brokerage, backed by Y Combinator, with offices in San Francisco and Atlanta. We do the insurance work end to end for about 1,000 new small businesses a month, with agents and people working the same accounts.
ROLE IN ONE LINE
Build and run the customer operation that takes Harper from thousands of small businesses to millions, with agents doing most of the work and people doing the work only people can.
Every policy we sell turns into a stream of work: getting it bound and active, payments and the finance agreements behind them, certificates and policy changes, cancellations to save, coverage to recover, and the hard cases that need an expert. Today that work runs through sixteen queues, worked by about forty people and more than a hundred bots that are only starting to take real volume. You make it one operation that scales.
WHY NOW
The volume is growing faster than headcount can, and it should. Our queues tell us exactly where the work is and where it waits: some move in hours, others build up for weeks because nobody owns them or the people who work them are spread across too many. We have built the machinery: queue channels, routing, status and assignment bots, task agents. What we need is the operator who decides who works what, sets the cadence, measures it by the hour, and moves work from people to agents one class at a time without dropping a customer.
WHAT YOU OWN
The operation. Every queue from first contact to active policy, renewal and recovery: arrivals, handle times, backlog, owners, service levels and cost per customer. You know every number by queue and by hour.
Capacity. Who works which queue, on which shift, in which office. Planning for 4x volume without 4x people, and for the day volume is ten times that.
The line between agents and people. Which work agents finish end to end, which a person approves, and which stays with a licensed expert. You move that line every week, on evidence.
Three centers. San Francisco for the people who build the system, Atlanta for the frontline and licensed teams, and our team in Pakistan for back-office work, each run by a lead who reports to you.
The talent model. A mix of domain experts (commercial lines, payments, compliance) and high-volume operators, plus an operator in every pod who builds and runs that pod’s agents.
Quality. One owner per account, honest answers to customers, and errors caught before they reach a carrier or a customer.
WHAT DONE MEANS
Every queue has one owner, a service level and a backlog that falls week over week.
Service people per thousand customers fall every quarter while retention rises.
At least two queues run with most items finished by agents, measured, with no rise in errors or complaints.
YOUR FIRST 90 DAYS
Day 30: you have worked the floor in all three locations, own the weekly operating review, and every queue has a named owner.
Day 60: capacity planning by queue and hour is running, and the oldest backlogs are falling.
Day 90: your plan for the next two quarters is written, with volume, agents and hires tied together and a path to millions of customers.
WHO THRIVES HERE
You have scaled customer operations or support to millions of customers or transactions, in delivery, marketplaces, consumer fintech, payments, insurance or SMB software at volume, and can show volume, headcount and cost per contact before and after.
You think in queues, arrivals, handle times and owners, and you have run workforce planning, tiered support, quality programs and offshore or partner teams.
You have moved real volume from people to automation or AI, know where it fails, and stayed accountable for the customer outcome while you did it.
You have led experts and high-volume operators in the same organization and know which work needs which.
You care deeply about customers, and the people who have worked for you would say so.
WHAT WE TEACH, WHAT YOU BRING
We teach commercial insurance and how Harper runs. You bring operations at scale, the judgment to pick which work goes to agents and which to people, and a record of building teams that got better as they grew. You do not need to be an engineer; you do need to use AI tools yourself.
WHO SHOULD NOT APPLY
Your experience is mainly SaaS customer success: strategic accounts, quarterly business reviews, renewals and solutions teams. That may matter here later; it is not this job now.
You manage through dashboards and never take a customer call or sit with a queue.
You scale by adding headcount first.
PAY, SCHEDULE AND PLACE
$220K to $260K, plus a 15 to 20% bonus and meaningful equity.
On site in San Francisco Monday to Friday, with a week in Atlanta most months and regular time with the team in Pakistan.
BENEFITS
Health, dental and vision. Breakfast, lunch and dinner at the office, commuter benefits and a gym membership.
HOW WE HIRE (USUALLY THREE TO FOUR WEEKS)
A conversation with the founders. We reply within 48 hours at every stage.
A working session on our real queue data: what you would change in the first 90 days, and what it takes to serve four times the customers.
Time with the teams in San Francisco and Atlanta, and a call with the Pakistan team.
References, then a decision within a week.
HOW TO APPLY
Send your resume and tell us about the largest customer operation you have run: how many customers or contacts, how many people, what you automated, and what the numbers looked like when you arrived and when you left.
Harper is an equal opportunity employer. If you need an accommodation at any point in the process, tell us and we will make it work. We will consider qualified applicants with arrest and conviction records in a manner consistent with the San Francisco Fair Chance Ordinance.
