Get a reply from this employer
Your resume and cover letter, tailored to exactly what they’re hiring for.
Chatporch is hiring a skilled UX writer and conversation designer to refine bot prompts and knowledge for hundreds of websites. You will read real conversations, spot where the bot stumbles, and rewrite prompts, tone, and rules to sound on-brand and helpful.
You’ll collaborate with engineers to turn patterns into product updates, build a playbook of best practices, and shape launch configurations for new sites. Remote role across the US with equity and a strong emphasis on clear, practical prose.
Chatporch runs the chat widget on hundreds of websites, and every one of them has its own voice, audience, and set of questions it needs to answer well. The system prompt for each site is assembled from a handful of slots: a description, a tone, a fallback instruction, free-form rules, and a knowledge blob. Picking the right words for those slots is the difference between a bot that feels like the business and a bot that feels like a chatbot.
That's the whole job.
Spend your mornings reading real conversations from `chatporch_conversations`. Look for the moments where the bot was stiff, wrong, vague, off-brand, or sent the visitor away when it shouldn't have. Then go fix the prompt slots, the knowledge blob, or the tone for that site.
You'll do this both reactively (a customer flags a bad transcript) and proactively (you sweep the top 20 sites by volume every two weeks). You'll build a private playbook of patterns we keep seeing: how to handle pricing questions for a service business, how to offer a callback without sounding pushy, what a roofing company's bot should say when someone asks about hail damage at 11pm.
When the patterns are strong enough, you'll work with engineering to turn them into product. New optional prompt slots. Better defaults. A library of starter configurations new customers can pick from. Eventually a fine-tuned eval set the AI engineer can score every model change against.
You write well. Not "writes good Slack messages" well. Actual prose that lands.
You've done one of: UX writing at a product company, conversation design at a voice or chat AI team, copywriting for service businesses, customer support leadership where you wrote macros and saw their open rates. Bonus if you've shipped a prompt for an LLM in production.
You're comfortable in a spreadsheet and you can read a SQL query someone else wrote. You don't need to write your own queries, but you do need to be willing to look at a hundred transcripts before forming an opinion.
You take feedback from real conversations more seriously than from your own taste. The job is to make the bot good for the visitor, not good for you.
You've worked in a service business (home services, dental, legal, real estate, fitness studios, restaurants). You speak Spanish or another second language. You have strong opinions about when a bot should hand off to a human and you can defend them.
$95k to $130k base depending on experience, plus early-stage equity. Fully remote within the US. Health, dental, vision, 4 weeks PTO. We pay for one writing conference or workshop a year.