Agent Builder
You will turn decades of expert knowledge into AI agents that sound like them, think like them, and meet their standards. This is editorial work at its core: deep reading, careful organization, and genuine respect for getting things right.
You will turn world-class experts into AI agents that think like them.
Onix is building Personal Intelligence: AI that belongs to you, protects your data, and helps you grow with guidance from real experts. We work with physicians, researchers, bestselling authors, and practitioners whose expertise reaches millions. Each has decades of knowledge: books, clinical protocols, podcast appearances, lectures, and proprietary frameworks. Your job is to turn all of that into an AI agent that sounds like them and meets their standards.
You will be the person who reads every book, watches every lecture, and organizes every protocol so that when a user talks to an expert’s AI agent, the response is accurate, nuanced, and unmistakably that expert’s voice. This is editorial work at its core. You are not writing marketing copy or managing social media. You are doing the deep, careful work of understanding how a world-class expert thinks and translating that into an AI experience that does them justice.
Think of this like being a research editor at a major publishing house or a fact-checker at a serious magazine. The difference between a good AI agent and a great one is the same difference between a good book and a great one: someone paid attention to the details.
What You Will Do
- Read and deeply understand each expert’s body of work: their perspective, voice, and areas of emphasis
- Collect, organize, and catalog expert content across formats (books, articles, research papers, transcripts, video, protocols)
- Work with engineering to build, test, and refine each expert’s AI agent
- Review agent outputs for accuracy, tone, and faithfulness to the expert’s actual positions
- Build and maintain content libraries with clear organization, tagging, and version tracking
Who You Are
You are the kind of person who reads a 400-page book and can explain its core argument in two paragraphs without losing the nuance. You have strong editorial instincts. You notice when a sentence sounds off, when a claim needs a source, when an AI-generated response drifts from what the expert would actually say. You are organized without being rigid, thorough without being slow, and curious enough that spending a week inside a physician’s body of work sounds like a good time rather than a chore.
You do not need a technical background. What you need is editorial judgment, intellectual stamina, and genuine respect for getting things right. You should already be using AI tools like ChatGPT and Claude in your daily work. Everyone here does.
You have:
- A degree in English, journalism, communications, publishing, or a related field (or equivalent demonstrated skill)
- Experience with editorial work: writing, editing, fact-checking, research, or content production
- Strong reading comprehension and the ability to synthesize complex material quickly
- Excellent written English with an ear for voice and tone
- Comfort working with large volumes of content across multiple formats
Qualifications
You are:
- A reader. You consume books, articles, and longform content naturally and you enjoy it
- Detail-oriented without being slow. You catch errors but you also hit deadlines
- Self-directed. You do not wait to be told what needs doing
- Already using AI tools daily. ChatGPT, Claude, or similar are part of how you work, not something you are curious about trying
- Curious about health, wellness, and human performance
- Comfortable working in person in Old Montreal
This Role is NOT For You If:
- You want to write original content or manage social media (this is editorial and curation work, not content creation)
- You skim rather than read
- You need detailed instructions for every task
- You want a remote job
- You see this as a stepping stone to a different function rather than a craft worth mastering
What Success Looks Like
- Every expert agent that goes live is accurate, well-sourced, and faithful to the expert’s voice
- Agents pass quality review on the first round because the preparation was thorough
- Content libraries are organized, tagged, and easy for any team member to navigate
- Experts themselves are impressed by how well the agent captures their thinking
- Quality improves with every agent you build because you refine the process each time