Harper is an AI-native commercial insurance company in San Francisco. We're not bolting AI onto insurance — we're rebuilding the entire business as software, on a simple bet: turning expert human judgment into compute is one of the largest transitions left to make, and a trillion-dollar industry still run 90% by hand is the place to prove it. We've grown ~100x in the last year and we move at that speed — on-site, in person, long days, very high standards. Almost no one joins Harper for insurance; they join to build the company that replaces how it works.
The role in one line
You're a senior commercial underwriter who thinks in systems — here to encode your judgment into the appetite rules, routing logic, and submission standards an AI can run, not just to quote and bind one account at a time.
Why this role exists now
Great underwriting judgment exists. Almost none of it compounds. It's trapped in individual inboxes, one-off decisions, messy submissions, and tribal knowledge — and that's most true in specialty commercial and E&S, where risk moves through emails, PDFs, ACORDs, supplementals, portals, spreadsheets, loss runs, unclear narratives, and broker notes that may or may not answer what an underwriter actually needs to know.
We're building the opposite: an operating system where underwriting judgment, submission quality, appetite logic, market routing, and broker workflows become infrastructure. This is exactly the "turn human judgment into compute" bet, pointed at the part of insurance that's hardest to systematize. Your judgment is the raw material.
This is not a traditional underwriting seat. You won't spend your day inside one carrier's guidelines saying yes or no. You'll help build what we collect, how we evaluate risk, how we package submissions, where we route accounts, when we escalate, and how we make Harper easier for markets to work with.
You don't need to write code. You do need to be the kind of underwriter who, when they see a bad submission, doesn't just think "this is incomplete" — they think: What should intake have asked? What should the system have extracted? What rule would have caught this? What makes this market-ready? What should we never send a carrier again? That instinct is the job.
What you'll do
Review live submissions and system outputs to catch missing context, weak risk summaries, bad assumptions, and incomplete exposure analysis.
Encode judgment into the product. Work directly with engineering and product to turn how you think into workflows, checklists, appetite rules, routing logic, escalation triggers, QA processes, and system instructions an AI can run.
Define "market-ready." Specify what a complete submission looks like across common commercial lines and specialty risks.
Build appetite and routing logic across carriers, wholesalers, MGAs, and specialty markets — what we write, where it goes, and why.
Advise across the company — intake, sales, placements, engineering, leadership — on what's placeable, what's missing, how to frame risk, and where workflows are breaking.
Write the playbooks for recurring risk categories and help the team understand how underwriting actually works in the real world.
Think like an underwriter and a broker at once: what's the risk, what will the market care about, how do we position it, what happens next?
Who you are
4+ years of commercial underwriting, with meaningful exposure to specialty commercial and/or E&S markets.
You've worked at an MGA, MGU, wholesaler, wholesale carrier, program administrator, specialty underwriting platform, or similar — and handled messy risks and imperfect submissions, not just clean admitted-market accounts.
Breadth across multiple commercial lines or risk categories — e.g. general liability, products liability, workers' comp, professional liability, tech E&O, cyber, commercial property, garage/garagekeepers, commercial auto, excess/umbrella, contractors, habitational, restaurants, staffing, security, healthcare, transportation, manufacturing, auto services, or other specialty classes.
Tech-savvy and product-curious. You don't code, but you're genuinely excited to sit with engineers, pressure-test system outputs, explain underwriting logic, and help build workflows that make the whole company smarter.
You explain underwriting clearly to engineers, salespeople, intake operators, brokers, and founders — without hiding behind jargon.
You move fast without perfect instructions, switch contexts easily, and you want your judgment to scale across thousands of submissions, not just the account in front of you.
Not the right fit
Someone who only wants to sit inside one carrier's appetite and apply guidelines.
Someone who wants a slow, traditional environment with clean handoffs and narrow responsibilities.
Someone who wants to advise from a distance but won't get into the workflow, the submission, the system output, the product logic, and the messy details.
The reality
On-site in San Francisco. The work touches intake, sales, placements, service, carrier relationships, product, and engineering — you'll operate with urgency, own outcomes, and help build the system while it's already running and quoting live business. There will be context switching, ambiguity, and problems that don't fit neatly in a job description. That's the point. The hours are long and the standards are high, because a rebuild this large doesn't happen part-time.
This is a builder role for an ambitious commercial underwriter who wants to step off the traditional track and help define what AI-native commercial insurance should look like. You've already seen what's broken. Come build what replaces it.
Compensation & benefits
Salary: $125,000–$175,000 + performance bonuses & equity
Location: San Francisco, in-office
Schedule: Monday–Friday, in-office hours that match the rest of the company. The hours are long.
Benefits: Uber commuter benefits; breakfast, lunch, and dinner provided; snacks/drinks/coffee daily; free gym membership; health, dental, and vision insurance.
Process
One to two screening calls — alignment on mission, pace, and role fit.
On-site super day — work a real submission, review system outputs, and meet product and engineering.