Generate Avatar Videos from Latest AI News using Dumpling AI and HeyGen
Overview
In a fast-paced digital world, staying updated with fresh content is key to visibility. This automation streamlines the process of discovering AI-related news, writing engaging commentary, and transforming it into avatar videos using HeyGen. Ideal for content creators, educators, and news publishers, this workflow eliminates the manual effort involved in research, script writing, and video creation.
By running this every hour, you ensure your brand shares timely, insightful AI updates through engaging videos—without ever opening an editing tool.
Full Workflow Breakdown
🕒 1. Schedule Trigger
- Node Type: Cron Trigger
- Purpose: Triggers the automation every hour.
- Why It Matters: Keeps your content pipeline active and autonomous.
2. Dumpling AI: Search for Recent AI News
- Node Type: HTTP Request (POST)
- Endpoint: https://app.dumplingai.com/api/v1/search-news
Payload:
{
“query”: “AI Agent”,
“language”: “en”,
“dateRange”: “pastHour”,
“page”: “1”
- }
- Purpose: Retrieves the latest AI news articles from trusted sources based on your query.
- Why It Matters: Ensures you’re sharing fresh and relevant news every hour.
3. Split Out Results
- Node Type: Split Out
- Purpose: Breaks the search results into individual news records.
- Why It Matters: Allows downstream modules to process each news article separately.
4. Limit to Top 4 Articles
- Node Type: Limit
- Purpose: Restricts the number of articles processed per cycle to four.
- Why It Matters: Keeps API costs under control and ensures processing speed.
5. Dumpling AI: Scrape Article Content
- Node Type: HTTP Request (POST)
- Endpoint: https://app.dumplingai.com/api/v1/scrape
Payload Template:
{
“url”: “{{ $json.link }}”,
“cleaned”: true
- }
- Purpose: Extracts clean, readable text from each article.
- Why It Matters: Gives GPT-4o rich, structured input to write coherent scripts.
6. Aggregate All Scraped Text
- Node Type: Merge
- Purpose: Combines all article texts into a single input.
- Why It Matters: Helps GPT-4o summarize multiple sources into one concise commentary.
7. GPT-4o: Generate Video Script
- Node Type: OpenAI Chat Model
- Model: gpt-4o
- Prompt Configuration:
- Instruct GPT to write a 30–60 second script summarizing the news.
- Must be single-line (no line breaks), simple language, and human tone.
- Why It Matters: Ensures the script works well with HeyGen’s avatar narration.
8. HeyGen: Generate Video
- Node Type: HTTP Request (POST)
- Endpoint: https://api.heygen.com/v2/video/generate
Payload Template:
{
“avatar_id”: “AVATAR_ID”,
“voice_id”: “VOICE_ID”,
“input_text”: “{{ $json.script }}”,
“video_size”: “1280×720”
- }
- Why It Matters: Converts AI-generated script into a professional avatar video.
9. Wait for Rendering
- Node Type: Wait
- Duration: 30 seconds
- Why It Matters: Gives HeyGen enough time to process video generation.
10. Check Video Status
- Node Type: HTTP Request (GET)
- Endpoint Template: https://api.heygen.com/v2/video/status?video_id={{ $json.video_id }}
- Why It Matters: Ensures the video is ready before proceeding.
11. If Not Ready, Retry
- Node Type: IF + Wait + Loop
- Condition: If status != completed
- Loop Wait: 20 seconds per retry
- Max Retries: 5
12. Log Video to Google Sheets
- Node Type: Google Sheets (Append Row)
- Content:
- Headline
- Source URL
- Generated video URL
- Why It Matters: Automatically logs finished videos for publishing or scheduling.
Real Use Case Example
- YouTube Shorts Channel: Auto-generate daily AI news summaries.
- LinkedIn Posts: Share trending updates using avatar videos.
- Internal Company Updates: Generate AI briefings for team members.
Conclusion
This workflow transforms raw AI headlines into polished avatar videos—on autopilot. It eliminates the bottleneck of manual research, writing, and production by leveraging Dumpling AI, GPT-4o, and HeyGen in perfect sync. You’ll always have fresh video content ready to publish across platforms.
