Editor, Agent Experience

Standard Template Labs • New York, NY

1d • $100,000 - $170,000

About The Position

STLabs is hiring an Editor, Agent Experience to own how its agent communicates with the people it helps. The agent resolves IT and HR requests in Slack and email for employees who may be stuck or irritated. This role sets the agent’s voice and defines the standard for a good interaction. Engineers build the models, prompts, and evaluation tooling; the editor determines the desired outcomes and provides examples of where the agent falls short.

The role involves maintaining behavioral guidance, reviewing real transcripts, translating findings into actionable requests for engineering, defining evaluation criteria, rewriting poor responses, building an exemplar library, and deciding when the agent should act, ask for more information, or route a request to a human. The product changes weekly, and the guidelines are expected to ship on a weekly cadence.

Requirements

• A strong ear for communication: notice when something is technically correct but lands badly, and explain why.
• Unusually strong plain-English writing for busy readers who may be frustrated; most output consists of written guidance and rewrites.
• Experience holding others to a standard across a high volume of real interactions, such as running a desk, editing a team, training staff, or grading conversations against a rubric.
• Ability to read thousands of transcripts, retain patterns, and turn cases into specific rules that are difficult to game.
• Comfort working with a product that changes weekly and adapting guidelines based on what breaks.
• A degree in literature, philosophy, or the humanities from a top school is required.

Nice To Haves

• ITSM experience is useful but not required.
• AI experience is useful but not required.
• Experience in hospitality, patient experience, editorial, teaching, support quality, or help-desk leadership is welcomed.

Responsibilities

• Write and maintain the agent’s voice and behavioral guidelines, including how it explains, asks questions, admits uncertainty, declines, escalates, and closes a loop.
• Review real transcripts every week and turn findings into precise, example-backed requests for engineering.
• Own the rubric for a good interaction and define what passes evaluation; engineers run the evaluations.
• Rewrite the agent’s worst answers and build an exemplar library for the team to learn from.
• Determine handoff behavior, including when the agent acts, asks, or stops and routes a request to a human.

What This Job Offers

• Job Type: Full-time