LinkedIn Voice Trainer & AI Content Generator

Train AI on your unique LinkedIn writing style, then generate authentic posts that sound exactly like you. Perfect for building a consistent personal brand while saving hours on content creation.

Why This Actor?

Writing authentic LinkedIn content takes time. This actor:

1. Analyzes your existing posts to understand your unique voice
2. Creates a reusable voice profile capturing your tone, style, and patterns
3. Generates new posts that sound exactly like you wrote them
4. Improves drafts to match your authentic style

Features

• Multi-Provider AI Support (BYOK): Use your own API keys for Claude, OpenAI, or Gemini
• Persistent Voice Profiles: Save and reuse voice profiles across runs
• Performance Analysis: Understand what makes YOUR posts successful
• Topic Suggestions: Get ideas that fit your voice and niche
• Draft Improvement: Enhance existing content to match your style

Tasks

1. analyze_voice - Create Voice Profile

Analyze your LinkedIn posts to extract your unique writing voice.

Input:

{
  "task": "analyze_voice",
  "aiProvider": "claude",
  "claudeApiKey": "sk-ant-...",
  "posts": [
    {
      "text": "Your post content...",
      "likes": 500,
      "comments": 50,
      "reposts": 20,
      "hashtags": ["leadership", "growth"],
      "contentType": "text"
    }
  ],
  "saveVoiceProfile": true,
  "voiceProfileName": "my-voice"
}

Output: Detailed voice profile with tone, style patterns, hook/CTA patterns, vocabulary preferences.

2. generate_post - Create New Posts

Generate LinkedIn posts in your voice.

Input:

{
  "task": "generate_post",
  "aiProvider": "claude",
  "claudeApiKey": "sk-ant-...",
  "voiceProfileId": "my-voice",
  "topic": "How I increased my team's productivity by 40%",
  "contentType": "text",
  "includeHashtags": true,
  "variations": 3
}

Output: Generated post matching your voice, plus alternative versions.

3. improve_draft - Enhance Existing Content

Make your draft sound more like "you."

Input:

{
  "task": "improve_draft",
  "aiProvider": "openai",
  "openaiApiKey": "sk-...",
  "voiceProfileId": "my-voice",
  "draft": "Here's my draft post that needs improvement...",
  "feedback": "Make it more engaging, add a question CTA"
}

4. suggest_topics - Get Content Ideas

Get topic ideas that fit your voice and niche.

Input:

{
  "task": "suggest_topics",
  "aiProvider": "gemini",
  "geminiApiKey": "...",
  "voiceProfileId": "my-voice",
  "industry": "SaaS",
  "recentTopics": ["productivity", "remote work"]
}

5. analyze_performance - Understand What Works

Analyze which patterns drive engagement for YOU.

Workflow: Training to Posting

Step 1: Scrape Your Posts

Use LinkedIn Profile Scraper to get your recent posts.

Step 2: Train Your Voice

{
  "task": "analyze_voice",
  "aiProvider": "claude",
  "claudeApiKey": "your-key",
  "posts": "{{ output from step 1 }}",
  "saveVoiceProfile": true,
  "voiceProfileName": "my-linkedin-voice"
}

Step 3: Generate Content

{
  "task": "generate_post",
  "voiceProfileId": "my-linkedin-voice",
  "topic": "Why I stopped using email for team communication"
}

Step 4: Post to LinkedIn

Use LinkedIn Poster to publish.

Voice Profile Storage

Voice profiles are stored in Apify Key-Value Store for reuse:

• Save: Set saveVoiceProfile: true and provide a voiceProfileName
• Load: Use voiceProfileId to load a previously saved profile
• Export: The profile is also in the actor output for external storage

AI Provider Comparison

Provider	Best For	Model Used
Claude	Complex analysis, nuanced writing	claude-sonnet-4-20250514
OpenAI	Fast generation, broad capability	gpt-4-turbo-preview
Gemini	Cost-effective, good quality	gemini-1.5-pro

Pricing

This actor uses pay-per-event pricing:

Event	Description	Price
apify-actor-start	Base cost per run	$0.10
voice_analyzed	Per voice profile created	$0.15
post_generated	Per post generated	$0.05
draft_improved	Per draft improved	$0.05
topics_suggested	Per topic suggestion set	$0.03

Example costs:

• Analyze voice + generate 5 posts: $0.10 + $0.15 + (5 × $0.05) = $0.50
• Generate 10 posts (existing voice): $0.10 + (10 × $0.05) = $0.60

Plus your AI provider's API costs

Tips for Best Results

1. Use 10-20 posts for voice analysis (more = better accuracy)
2. Include high-performing posts - they represent your best voice
3. Re-train monthly as your style evolves
4. Combine with scrapers for automated workflows

Integration with Zapier/Make/n8n

This actor works great in automation workflows:

1. Trigger: Weekly schedule
2. Action 1: Run linkedin-profile-scraper for new posts
3. Action 2: Run linkedin-voice-trainer to generate content
4. Action 3: Send to approval (Slack, email, etc.)
5. Action 4: On approval, post via linkedin-poster

LinkedIn Authentication

When using linkedinProfileUrl to automatically scrape posts, you'll need to provide LinkedIn credentials:

Option 1: Email/Password (Simple, but MFA must be disabled)

Provide linkedinEmail and linkedinPassword in the input. Works for accounts without 2FA/MFA.

Option 2: Cookies (Recommended, works with MFA)

If you have Multi-Factor Authentication (MFA/2FA) enabled, you must use cookies.

Step 1: Install a cookie export extension:

• Cookie-Editor by Moustachauve - Clean interface, supports Chrome/Firefox/Edge/Safari
• EditThisCookie by Francesco Capano - Chrome extension, classic choice

Step 2: Log into LinkedIn in your browser

Step 3: Click the cookie extension icon, then:

• Cookie-Editor: Click "Export" → "Export as JSON"
• EditThisCookie: Click the export button (looks like ↓)

Step 4: Paste the JSON array when running the LinkedIn Profile Scraper

The most important cookie is li_at - this is your LinkedIn session token.

Support

• Documentation
• Issues

---

Built by John Rippy | johnrippy.link

---

Keywords

linkedin voice trainer, ai writing assistant, content personalization, brand voice ai, linkedin content ai, openai integration, claude integration, gemini integration