Build a voice-trained LinkedIn content engine that does not sound like AI

Gather 10-20 source samples from the person whose voice you want to preserve: LinkedIn posts, emails, podcast clips, sales call excerpts, internal rants, and rough notes. Work in descript and airtable and capture the concrete operating details: source asset, audience, hook, angle, channel, draft copy, visual asset, approval status, and publish URL. Keep the work narrow: complete step 1, record the decision or asset created, and avoid changing unrelated parts of the workflow. Before moving on, check that the asset matches the channel, the CTA is clear, and the final version is easy to reuse; the expected handoff is A source library of writing and speech samples tagged by topic and tone.

Use Claude to analyze the source library and produce a voice profile with sentence patterns, recurring beliefs, strong opinions, favorite phrases, banned phrases, pacing, humor level, and examples of acceptable and unacceptable rewrites. Work in claude and airtable and capture the concrete operating details: source asset, audience, hook, angle, channel, draft copy, visual asset, approval status, and publish URL. When you use the prompt template attached to this step, paste the real source material, name the expected output sections, and ask the model to flag uncertainty instead of filling gaps. Before moving on, check that the asset matches the channel, the CTA is clear, and the final version is easy to reuse; the expected handoff is A reusable voice profile and banned phrase list for future drafts.

Create an Airtable table for raw ideas with columns for source, rough thought, audience, pain point, proof, story, format, priority, draft status, reviewer, and performance notes. Work in airtable and capture the concrete operating details: source asset, audience, hook, angle, channel, draft copy, visual asset, approval status, and publish URL. Keep the work narrow: complete step 3, record the decision or asset created, and avoid changing unrelated parts of the workflow. Before moving on, check that the asset matches the channel, the CTA is clear, and the final version is easy to reuse; the expected handoff is A repeatable idea intake table that prevents LinkedIn content from depending on blank-page brainstorming.

Use Claude to convert raw ideas into post briefs, not final posts yet. Work in claude and airtable and capture the concrete operating details: source links, buyer context, audience, claims, proof points, owner, review status, and the final deliverable link. When you use the prompt template attached to this step, paste the real source material, name the expected output sections, and ask the model to flag uncertainty instead of filling gaps. Before moving on, check that the asset matches the channel, the CTA is clear, and the final version is easy to reuse; the expected handoff is A weekly queue of LinkedIn post briefs ready for drafting.

Use Claude to draft one sharp version and one safer version for each approved brief. Work in claude and capture the concrete operating details: source asset, audience, hook, angle, channel, draft copy, visual asset, approval status, and publish URL. When you use the prompt template attached to this step, paste the real source material, name the expected output sections, and ask the model to flag uncertainty instead of filling gaps. Before moving on, check that the prompt includes the actual inputs, the requested output format, and clear rules against invented facts; the expected handoff is Draft LinkedIn posts that follow the voice profile instead of sounding like generic AI output.

Before scheduling, run every draft through a QA pass. Work in claude and authoredup and capture the concrete operating details: source asset, audience, hook, angle, channel, draft copy, visual asset, approval status, and publish URL. When you use the prompt template attached to this step, paste the real source material, name the expected output sections, and ask the model to flag uncertainty instead of filling gaps. Before moving on, check that the asset matches the channel, the CTA is clear, and the final version is easy to reuse; the expected handoff is Approved drafts with formatting, hook, and voice issues corrected.

Schedule approved posts and track performance in Shield. Work in shield and airtable and capture the concrete operating details: source asset, audience, hook, angle, channel, draft copy, visual asset, approval status, and publish URL. Keep the work narrow: complete step 7, record the decision or asset created, and avoid changing unrelated parts of the workflow. Before moving on, check that the asset matches the channel, the CTA is clear, and the final version is easy to reuse; the expected handoff is A performance feedback loop that improves the next month of posts.