Turn data into revenue. Build the analytics that power growth for the best AI commercial insurance distribution engine.

The Problem

36 million businesses in America need insurance. It’s not optional—regulations require it, contracts demand it, landlords won’t lease without it. Yet the industry is broken.

77% of small businesses are underinsured. 40% have no coverage at all. They’re running with risk because the distribution system failed them—too slow, too opaque, too confusing to navigate.

Over 90% of commercial insurance distribution is still human-led. We’re building the inverse: 90%+ AI-led, inching toward the higher 90s.

To grow that fast, we need to understand—with precision—what’s working, what’s not, and why.

The Thesis

Every industry with human-bounded distribution consolidates rapidly once someone makes it computational. When distribution becomes computational, Jevons Paradox kicks in: efficiency leads to expansion.

Insurance will follow this pattern. When getting coverage becomes fast and frictionless, the 77% of underinsured businesses will finally get properly protected. The market expands, not contracts.

We’re building the engine that makes that happen. You’ll build the intelligence that tells us how well it’s working.

The Role

You’ll own the analytics that power growth, sales, and marketing decisions. This isn’t a “build dashboards and wait for questions” role. You’ll work directly with founders and growth leads to identify opportunities, run experiments, and prove what works.

We have data other companies don’t. Every customer conversation—transcribed, analyzed, searchable. Every AI decision—what context it had, what it chose, what happened next. Every sales touchpoint across voice, email, and web. Millions of decision traces that reveal what actually drives conversion.

What makes this different: You’re not supporting a sales team—you’re building the intelligence that makes an AI-native sales motion work. Your models will directly power how our agents prioritize leads, time outreach, and personalize conversations. The line between “analytics” and “product” barely exists here.

We’re a growing team looking for high-agency operators who can turn data into decisions.

What You’ll Build

Lead intelligence. Which leads convert? Which waste time? Build models that score prospects based on signals others don’t have—conversation sentiment, engagement patterns, business characteristics. Your models will directly determine which leads our agents call first.

Conversation analytics. Thousands of AI-powered sales and service conversations happening daily. What patterns predict close? What objections kill deals? What phrases build trust? You’ll mine this data and feed insights back into our agents.

Attribution that actually works. Multi-touch attribution across voice, email, web, and referral channels. Not “last click wins” attribution—real causal inference on what drives revenue.

Agent evaluation. Is the AI getting better? Build the systems that measure agent performance against business outcomes—not just “did it follow the script” but “did it close the deal.”

Experiment infrastructure. A/B testing frameworks that let us iterate on agents, pricing, and campaigns with statistical rigor. The feedback loops that make everything compound.

You Might Be a Fit If…

You’re a full-stack data scientist. You pull your own data, build your own models, deploy your own solutions. You don’t wait for data engineering to unblock you.

You’ve worked on GTM problems. Lead scoring, attribution, sales analytics—you’ve built models that drive revenue, not just research.

You think in experiments. You design tests, measure results, and iterate. You know the difference between correlation and causation.

You can communicate to non-technical stakeholders. You can explain to the CEO why a metric moved and what to do next.

You’re comfortable with ambiguity. We don’t have a data warehouse with perfect schemas. You’ll figure out what data we have and how to use it.

You’re 2-5 years into your career. Enough experience to be dangerous, not so senior you’ve forgotten how to write SQL.

Compensation

Salary: $130,000 - $190,000

Equity: 0.05% - 0.25%

Location: San Francisco, in-office. We build together.

The Process

1. 15-min founder call — Alignment on mission and pace

2. Technical conversation — Walk us through analysis you’ve done

3. On-site — Meet the team, see the data

To Apply

We’re building a vertically integrated AI platform that connects go-to-market, sales operations, customer service, and retention under one architectural roof. That integration creates compounding through feedback loops—every interaction makes the system smarter. Thousands of businesses already trust us.

Most data science roles are removed from business impact. You build models, hand them off, and hope someone uses them. At Harper, you’ll work directly with founders and growth leads. Your analyses will shape strategy. Your models will power production systems. Your experiments will directly move revenue.

Data talks. Narratives walk. If you prove things instead of just believing them—send your resume and an example of analysis that drove a business decision.

We’re a championship-minded team. We push each other. We move fast. We care about craft. If that sounds like where you belong, let’s talk.