Model Behavior Specialist

Hark • San Jose, CA

6d • $150,000 - $250,000

About The Position

Hark is hiring a hands-on, high-ownership individual contributor to own the quality of the customer experience with its personal AI assistant. The specialist will test Hark end to end on everyday tasks, identify shortcomings, trace failures to their root causes, and drive improvements with engineering and design. The role is part of a small, new team, and its work will be directly visible in the product.

The specialist will create and refine guidelines for how Hark handles real-world tasks, define quality standards, measure the effects of product changes, benchmark Hark against other AI products, grade AI outputs, build test sets, and recommend whether new versions are ready to ship.

Requirements

• At least 2 years of experience owning demanding, high-stakes work end to end in any field.
• Strong attention to detail and the ability to distinguish a good experience from a great one.
• Clear written and oral communication skills, including the ability to defend recommendations in writing and verbally.
• Exceptional root-cause analysis skills and the ability to drive fixes beyond addressing symptoms.
• Strong data-driven decision-making skills, including using data to assess whether changes worked and acting on the results.
• Self-starter who takes ownership of outcomes and remains rigorous through high-volume, iterative work.

Nice To Haves

• Experience with prompt engineering and evaluations.
• Background in management consulting or investment banking.
• Experience in e-commerce, payments, travel, or other consumer transactions.

Responsibilities

• Write, test, and continuously refine product guidelines for how Hark handles real-world tasks.
• Define what constitutes a great experience for each task and hold the product to that quality bar.
• Test Hark end to end using real tasks, identify shortcomings, and trace failures to their root causes.
• Drive identified issues to resolution with engineering and design, then verify that fixes work.
• Benchmark Hark against other AI products and use data to measure how product changes affect the customer experience.
• Grade AI outputs, build test sets, and recommend whether new versions are ready to ship.

What This Job Offers

• Job Type: Full-time
• Career Level: Mid Level
• Education Level: No Education Listed