Product Engineer

Harper
San Francisco, CA
On-site

Who this role is best for

Geared toward candidates comfortable with rapid AI deployment and hands-on collaboration, this role suits those with production AI experience and a preference for on-site work in San Francisco.

Best fit for

  • Candidates with experience shipping AI applications and a preference for direct, on-site collaboration with business teams
    — “Your code does not sit in a repo. It runs the business the same week you write it.”
  • Individuals who prioritize business impact over code polish and can quickly adapt to new AI models
    — “When a new model or capability lands, you test it that day and ship it if it helps.”
  • Engineers who can build and operate full-stack AI systems and are driven by measurable outcomes
    — “Ship AI features end to end. Frontend, backend, agents, and evals. You own the whole thing, not a slice.”

Things to consider

  • This role demands full-time on-site presence with no remote or hybrid flexibility
    — “This is not the seat if you want specs handed to you, if you would rather polish code than talk to the people using it, or if you need a remote or hybrid schedule.”
  • Candidates must be ready to take full ownership of features and their business impact, even if the feature does not perform as expected
    — “the outcome is yours, including when the feature you loved does not move the number.”

How to stand out

  • Highlight projects where your AI work directly influenced business metrics in your resume and application
    — “Send your resume and a link to something you built that had measurable business impact.”
  • Demonstrate your ability to work across the stack and explain technical outcomes to non-technical stakeholders
    — “You work across the stack in Python, TypeScript, or similar, and you write code with AI tools like Cursor or Claude Code.”
  • Emphasize your experience with AI tooling and how it accelerated your development and deployment cycles
    — “Move fast with AI tooling. Run several coding sessions at once and compress weeks into days.”
  • Show how you identify problems before they are reported and how you own the solution from start to finish
    — “find what is broken before they tell you, and own the outcome.”
  • Include examples of how you built systems that improve with each interaction and compound over time
    — “Build systems that compound. Trace every decision and feed every outcome back, so the AI gets better with each interaction.”
Pace · Fast PacedCollaboration · HighAutonomy · HighDecision Impact · Team

Derived from job-description analysis by Serendipath's career intelligence engine.

What success looks like

  • Ship AI features end to end
  • Embed with operators
  • Build systems that compound
  • Stay on the frontier
  • Move fast with AI tooling
Typical background
Experience shipping AI to productionExperience working across the stack in Python, TypeScript, or similar

Skills & requirements

Required

AI Feature DevelopmentPythonTypeScriptAI ToolingSystem Design

Preferred

LLM ApplicationsAgent PipelinesWorkflow AutomationVoice AIRAGEval SystemsOperations Tooling

Stack & domain

PythonTypeScriptLLM ApplicationsAgent PipelinesWorkflow AutomationAI ToolingCursorClaude CodeLeadershipCommunicationProblem-solvingAICommercial InsuranceHealthcare

About the role

Original posting from Harper via Ashby

ABOUT HARPER

Harper is an AI-native commercial insurance brokerage for small businesses. We built our own software for nearly every step, so our people spend their time where trust matters. We use AI to earn that trust.

Your code does not sit in a repo. It runs the business the same week you write it.

On Monday you sit with the sales floor and watch account executives retype the same details three times. On Tuesday you ship an agent that fills them in. By Wednesday you can show that calls got shorter and more of them closed. Then you go find the next one.

At most companies, engineers get a ticket and never meet the person who filed it. Here you embed with the operators in sales, service, and underwriting, find what is broken before they tell you, and own the outcome.

WHAT YOU'LL DO

  • Ship AI features end to end. Frontend, backend, agents, and evals. You own the whole thing, not a slice.
  • Embed with operators. Sit with sales, service, and underwriting, and find what is broken before they tell you.
  • Build systems that compound. Trace every decision and feed every outcome back, so the AI gets better with each interaction.
  • Stay on the frontier. When a new model or capability lands, you test it that day and ship it if it helps.
  • Move fast with AI tooling. Run several coding sessions at once and compress weeks into days.
  • Own the metric. Set up the measurement, track the result, and show the company what moved.

WHAT YOU NEED

  • You have shipped AI to production: LLM applications, agent pipelines, or workflow automation serving real users.
  • You work across the stack in Python, TypeScript, or similar, and you write code with AI tools like Cursor or Claude Code.
  • You care more about user and business impact than code elegance, and you can explain to non-technical people why a number moved.

Voice AI, RAG, eval systems, or operations tooling experience helps, and it is not required. We hire across levels and set the level during interviews. If you meet most of this, apply.

THE REALITY: READ THIS BEFORE YOU APPLY

This job is measured by what moved, not what shipped. You work directly with founders, with no committee and no approval chain. That means the outcome is yours, including when the feature you loved does not move the number.

This is not the seat if you want specs handed to you, if you would rather polish code than talk to the people using it, or if you need a remote or hybrid schedule.

It is the seat if writing code that runs a real business that same week sounds like the work you want.

Harper is an equal opportunity employer.

HOW WE HIRE

  • A 60-minute technical screen: a deep dive on a past project and a system design.
  • A conversation with an engineering lead on fit and how you work.
  • A super day in the office, sitting in on the operation and doing real work alongside us.

THE BASICS

  • Pay: $140,000 to $280,000 base, depending on level. Bonus and equity eligible.
  • Where: On-site, five days a week, in the room with the team. San Francisco at 425 Market Street.
  • Hours: Very early morning start, 8am to 6pm PT, Monday through Friday.
  • Benefits: Health, dental, and vision. San Francisco adds three meals a day in the office, commuter benefits, and a gym membership.

HOW TO APPLY

Send your resume and a link to something you built that had measurable business impact. Tell us the problem you found, what you shipped, and the number it moved.

Source: Harper careers (Ashby)

Similar roles

More roles at Harper