AI Content & Voice Expert — Help Us Build a Client Voice System

We’re looking for someone who deeply understands how to leverage AI—especially tools like Claude—to help us design a better way to create content in the authentic voice of our clients. We create thought-leadership and personal-brand content for founders, experts, and executives. The goal is to build a system where we can feed AI a large body of information about a client—transcripts, interviews, books, podcasts, existing content, strategy calls, approved scripts, feedback, etc.—and have it develop a deep understanding of: * How they naturally speak and structure thoughts * Their vocabulary and language patterns * Their beliefs, philosophies, and opinions * Their stories, frameworks, and expertise * What they would and wouldn’t naturally say Think of it as creating a **“voice and thinking fingerprint”** for each client. We want AI to help our writers create content that feels like **the client working with an exceptional writer who understands them extremely well**—not generic AI writing. We already have developers, so this is **not primarily a software engineering role.** We need someone who can help us figure out the right system and methodology: * How client knowledge should be organized * How to create and maintain a reliable voice profile * How Claude or other AI tools should access and use that information * How to prevent generic AI writing and hallucinated opinions/stories * How writers should interact with the system * How client edits and feedback can continuously improve future output * Whether this should live in Claude Projects, use APIs/RAG, or become a custom internal tool You should be highly experienced with modern AI tools, prompt/context design, knowledge systems, and AI-assisted writing workflows. You do **not** necessarily need to be a developer, but you should understand the technology well enough to work with our development team. ## When Applying Please answer these questions: 1. How would you approach creating a reliable “voice and thinking fingerprint” from hundreds of hours of a person’s content? 2. Would you start with Claude Projects or build something more custom? Why? 3. How would you prevent the output from falling into recognizable generic AI writing patterns? 4. How would you make sure the AI distinguishes between something the client actually believes and something it is inventing or inferring? 5. Have you built anything similar? If so, please share examples. We care much more about **how you think about this problem** than AI buzzwords.