User onboarding video series from feature documentation

Creating user onboarding videos is one of those tasks that sounds straightforward until you're actually doing it. You need to convert your feature documentation into engaging video content, summarise the key points so viewers don't get lost, and produce it at scale without spending weeks in post-production. Most teams solve this by hiring a video production agency or assigning someone to manually create each video; both approaches drain budgets and timelines.

What if you could automate the entire process? Take your existing feature documentation, distil it into a concise script, generate a professional video with a digital presenter, and have the final output ready for your onboarding flow without touching it once after you kick things off. This workflow combines three powerful AI tools that each do one thing well: HourOne for video generation with a presenter, Resoomer AI for intelligent summarisation, and Synthesia for additional video polish and customisation.

This is an intermediate-level workflow because you'll need to understand API authentication, payload structures, and basic conditional logic in your orchestration tool. If you've set up a Zapier or Make automation before, you'll find this straightforward. The payoff is significant: you can produce a complete onboarding video series in the time it would normally take to write one script.

The Automated Workflow

We'll build this workflow using Make (formerly Integromat) as the orchestration layer, though the same logic works in Zapier or n8n with minor adjustments. The workflow follows this sequence:

1. Documentation text enters the system (from a form, email, or API call).
2. Resoomer AI summarises the documentation into a tight script.
3. Synthesia generates a video with visual elements based on the script.
4. HourOne polishes the output with a professional digital presenter.
5. The final video is saved and a notification is sent.

Setting Up Your Make Scenario

Create a new scenario in Make and start with an HTTP webhook trigger. This allows you to POST documentation to the workflow from anywhere.

Copy
POST https://hook.make.com/handshake/[your-scenario-id]
Content-Type: application/json

{
  "documentation_text": "Your feature documentation here",
  "feature_name": "User Dashboard",
  "target_audience": "New users"
}

Make will generate a unique webhook URL for your scenario. Test it once, then save it; you'll use this endpoint whenever you want to generate a new onboarding video.

Step 1: Summarise with Resoomer AI

Resoomer's API takes long-form text and extracts the most important points. This creates a tight script suitable for video narration; you're not trying to cram everything into 90 seconds, but 300-400 words of focused content performs much better than 2000 words of documentation read aloud.

First, obtain your Resoomer API key from your account settings. In Make, add an HTTP request module and configure it like this:

Copy
POST https://api.resoomer.com/api/summarize
Content-Type: application/json
Authorization: Bearer [YOUR_RESOOMER_API_KEY]

{
  "text": "{{1.documentation_text}}",
  "type": 2,
  "language": "en",
  "summarize_to": 30
}

The summarize_to parameter sets the summary length as a percentage of the original. Setting it to 30 means Resoomer will create a summary that's roughly 30% of the input length. For a 2000-word documentation file, this gives you around 600 words, which is ideal for a 3 to 5-minute video.

Map the output to a variable called summary_text. You'll use this in the next steps.

Step 2: Generate the Video with Synthesia

Synthesia's API lets you create videos with digital presenters from text. It's particularly good for consistent, professional-looking output. You'll need a Synthesia account and API key.

Add another HTTP request module in Make:

Copy
POST https://api.synthesia.io/v1/videos
Content-Type: application/json
Authorization: Bearer [YOUR_SYNTHESIA_API_KEY]

{
  "input": [
    {
      "type": "speech",
      "text": "{{summary_text}}",
      "avatar": "amy",
      "background": {
        "type": "color",
        "color": "#FFFFFF"
      }
    }
  ],
  "test": false,
  "title": "Onboarding: {{1.feature_name}}"
}

This creates a video with Amy (one of Synthesia's digital presenters) reading your summarised script against a white background. The API returns a video_id immediately, but the actual video generation happens asynchronously. Save the video_id in a variable for polling later.

If you want a more branded aesthetic, swap the avatar value (try "oliver", "maya", or "john") and adjust the background colour to match your brand guidelines.

Step 3: Poll for Video Completion

Synthesia doesn't generate videos instantly; they typically finish within 2 to 5 minutes. You need to poll the API to check when it's done. Add a repeat loop in Make and call the status endpoint:

Copy
GET https://api.synthesia.io/v1/videos/{{video_id}}
Authorization: Bearer [YOUR_SYNTHESIA_API_KEY]

The response includes a status field. Keep polling until it reads "COMPLETED" (or "FAILED", which indicates an error). Set the loop to check every 30 seconds and time out after 10 minutes. Once complete, the response includes a download_url where you can download the MP4 file.

Step 4: Enhance with HourOne (Optional)

HourOne excels at generating videos with realistic human presenters. If you want additional polish or a human-like presenter rather than a digital avatar, route the Synthesia output to HourOne. However, this adds complexity, so only use it if your onboarding videos need that extra production quality.

To use HourOne, you'll submit the summarised script:

Copy
POST https://api.hourone.ai/api/videos/create
Content-Type: application/json
Authorization: Bearer [YOUR_HOURONE_API_KEY]

{
  "script": "{{summary_text}}",
  "presenter": "emma",
  "video_quality": "1080p",
  "output_format": "mp4"
}

HourOne returns a jobId. Poll its status endpoint until the video is ready, much like Synthesia. The advantage is that HourOne's output often looks more polished and presentation-like, though it does add 5 to 10 minutes to the overall process.

In most cases, skip this step and use only Synthesia; you'll get 80% of the quality in half the time.

Step 5: Upload and Notify

Once your video is complete, you have several options:

• Save it to cloud storage (Google Drive, Dropbox, AWS S3) so team members can access it immediately.

• Push it to your learning management system or onboarding platform via their API.

• Email the download link to stakeholders.

In Make, add a Google Drive upload module or a webhook call to your backend:

Copy
POST https://your-domain.com/api/onboarding-videos
Content-Type: application/json
Authorization: Bearer [YOUR_API_KEY]

{
  "feature_name": "{{1.feature_name}}",
  "video_url": "{{3.download_url}}",
  "generated_at": "{{now}}",
  "summary": "{{summary_text}}"
}

This stores metadata about the video in your database, making it searchable and trackable later.

Complete Workflow Summary

To recap, the flow looks like this:

1. Documentation enters via webhook.
2. Resoomer summarises it.
3. Synthesia generates a video from the summary.
4. Make polls Synthesia until the video is ready.
5. The video is uploaded to cloud storage or your backend.
6. A Slack notification or email confirms completion.

Total execution time: 5 to 10 minutes from start to finish, depending on video length and system load. You can batch multiple videos by triggering the webhook multiple times in succession.

The Manual Alternative

If you want more control over individual steps, you can run each tool separately. Download your documentation, paste it into Resoomer's web interface and manually refine the summary. Copy the summary into Synthesia's editor, tweak the avatar and background, and generate the video by hand. Download the MP4 when it's ready.

This approach takes 15 to 20 minutes per video and gives you granular control over phrasing, presenter tone, and visual styling. Use it when you're creating your first few videos and want to perfect the tone and appearance before automating. Once you've settled on a style, automation becomes worthwhile.

Pro Tips

1. Refine Resoomer's Output for Better Results

Resoomer creates a good starting point, but AI-generated summaries sometimes lose important context or include awkward phrasing. Before sending the summary to Synthesia, add a quick manual review step. Use Make's "Wait for Webhook" module to pause the workflow and notify yourself (via Slack or email) to review the summary. Approve or edit it, then the workflow resumes. This prevents you from generating videos with confusing scripts.

2. Watch Your API Rate Limits

Resoomer allows 100 API calls per day on most plans. Synthesia offers 60 videos per month on their Starter plan. HourOne is similarly rate-limited. If you're generating dozens of onboarding videos in a single day, stagger your requests using Make's scheduler, spreading submissions across several hours. Check each tool's API documentation for exact limits before launching automation.

3. Test with a Short Documentation Sample First

Run your workflow with a small, 200-word documentation excerpt before automating dozens of videos. This lets you catch configuration errors, check the quality of generated output, and confirm that videos are being saved to the right location. Debugging a workflow while it's generating 50 videos in parallel is far more painful than fixing issues upfront.

4. Use Webhooks to Trigger from Your Content Management System

If your feature documentation lives in a wiki, knowledge base, or CMS, add a webhook call to your automation whenever new documentation is published. Zapier, n8n, and Make all support incoming webhooks from third-party services. This way, new onboarding videos are automatically generated minutes after documentation is written, keeping them in sync.

5. Customise Synthesia's Branding Elements

Synthesia allows you to set background colours, add logos, and customise the presenter's appearance. Hardcode your brand colour and logo into your Make scenario so every generated video is consistent. This only takes a moment in the initial setup but saves you from having to edit videos manually later.

Cost Breakdown

For a small team creating 10 to 20 onboarding videos per month, the base setup (Resoomer + Synthesia + Make) is cost-effective. You're paying roughly £0.60 per video in tool fees, which is a fraction of the cost of hiring someone to create them manually.

If you need to produce more than 60 videos monthly, upgrade Synthesia to their Pro plan. If quality is paramount and you're willing to pay more, add HourOne for the best-looking output. Most teams find the Synthesia-only approach strikes the right balance between cost, quality, and speed.

Building this workflow takes about an hour to set up the first time. The payoff comes immediately: subsequent videos are generated in minutes without any human intervention beyond submitting the documentation. After a month of running this automation, you'll have created more onboarding videos than most teams produce in six months, and you'll have spent a fraction of what hiring someone full-time would cost.