Turn long show transcripts into clear, useful email recaps for your team. The flow builds a short summary, extracts discussion topics with smart questions, researches each topic, and sends everything to Gmail as a clean digest. It suits content teams, podcast hosts, and marketers who need quick insights without manual note taking.

Here is how it runs. You click to start. A transcript is loaded, then split into chunks so the model can handle long text. An OpenAI model writes a concise summary. A second step extracts topics and follow up questions using a fixed schema. Each question is split into separate items and sent to an AI Agent that uses Wikipedia to explain the topic in simple terms. The results are formatted as HTML and emailed through Gmail. You get a ready to read digest that supports content planning and research.

Setup is simple. Add your OpenAI API key and connect Gmail with OAuth. Paste or generate your transcript, adjust chunk size and prompts if needed, and set the recipient email. Expect faster prep time, reliable structure, and consistent talking points. Use it for show notes, blog drafts, or weekly team updates.

• Manual start lets you control when to process a new transcript
• Loads and splits long text using a recursive text splitter for better model handling
• Summarizes the transcript with an OpenAI chat model for a concise recap
• Extracts topics and follow up questions using a defined JSON schema
• Splits questions into separate items to process each topic on its own
• AI Agent researches topics using the Wikipedia tool for simple explanations
• Formats titles, topics, and summary into clean HTML with a code node
• Sends a final digest to Gmail with subject and body ready for stakeholders

• Reduce manual research from 3 hours to 15 minutes per episode
• Automate about 80% of topic exploration with AI and Wikipedia
• Improve summary consistency and structure across all episodes
• Handle long transcripts reliably by splitting content into chunks
• Connect OpenAI and Gmail so insights are delivered in your inbox

1. Import the template into n8n: Create a new workflow in n8n > Click the three dots menu > Select 'Import from File' > Choose the downloaded JSON file.
2. You'll need accounts with OpenAI and Gmail. See the Tools Required section above for links to create accounts with these services.
3. In n8n, open the OpenAI Chat Model nodes, choose Credential to connect with, click Create new credential, and enter your OpenAI API key from the OpenAI account API page.
4. Open the Gmail node, choose Credential to connect with, click Create new credential, and complete the OAuth flow for your Google account. Select the Gmail account you want to send from.
5. Open the Podcast Episode Transcript code node and paste your episode transcript into the transcript field. Save the node.
6. Check the Recursive Character Text Splitter settings. Keep chunk size near 6000 and overlap near 1000 for long content, or adjust based on your model limits.
7. Review the Summarize Transcript node and confirm it uses the OpenAI credential you set. Keep operation mode as document loader so it ingests the split text correctly.
8. Open the Extract Topics and Questions node and confirm the schema fields match your needs. Keep questions and topics as arrays for clean formatting.
9. Open the AI Agent node. Ensure the Wikipedia tool is connected and the agent uses your OpenAI credential. Adjust the prompt text to match your brand voice if needed.
10. Edit the Format topic text and title code node if you want different HTML structure or headings.
11. Open the Gmail node and set the To field to your target inbox. Update the subject line if you want a different naming style.
12. Click Execute Workflow to run a test. Confirm the summary, topics, questions, and research appear in the email. If the email is missing content, check that the splitter, summarizer, and extractor nodes each output data and that credentials are connected.