Build Your First AI App from Scratch

No dark-mode styles — this email renders the same in a dark client.

From *'I Have an Idea'* to 
===========================

*'I Have a Working Workflow'*
=============================


The visual canvas can feel unfamiliar at first. No worries, we're going to build your very own AI app completely from scratch today. Also, pick up basic prompting skill that makes it actually useful along the way.


We're going to build a Meeting Minutes Summarizer which is an app that can read meeting transcripts and extract action items and key decisions for you.


**Create Your ****Meeting Minutes Summarizer in 3 Minutes**


1. **Start fresh:** Open Studio, click **Create from Blank**, then select **Workflow**.

2. **Set up the input (User Input):** Choose User Input as the start node. 
What information does this app need to begin? Add your input variables here. You can think of variables as simple placeholders for the actual text or files you will provide later when running the app (e.g., create a variable named meeting_transcript and set its type to Paragraph).

3. **Process the data (LLM Node):** What should the AI do with that information? Connect an LLM Node to the Start Node. Write your System prompt and reference your variables using the {variable_name} syntax. A good prompt is specific, giving it a **role**, a clear **goal**, and **guides** to follow. 


4. **Capture the result (Output Node): **What should the AI produce? Connect an End Node to the LLM and map the LLM's text output to an output variable to capture your structured summary. If you skip this step, the output will be blank.

5. **Test it:** Don't worry about making it perfect. Simply click **Test Run**(top right) to see the generated results. Edit the the prompt or switch to different LLMs for comparison.


[ **Build Now**](https://cloud.dify.ai/apps?utm_medium=email&utm_source=brevo&utm_campaign=user_onboarding&utm_content=email_2 "Button title")


**Helpful Resources**


- [Get to Know Variables](https://docs.dify.ai/en/use-dify/getting-started/key-concepts#variables) 
- [Understand LLM Tags](https://docs.dify.ai/en/use-dify/tutorials/workflow-101/lesson-03#understand-the-tags) 
- 

[How to Write LLM-Ready Prompt](https://dify.ai/blog/prompt-engineering-for-workflow-ready-llm-apps)


We’re here to make your experience amazing. If you have any questions or feedback, feel free to reply to this email, we’d love to hear from you!


LangGenius, Inc.

101 Jefferson Drive, Menlo Park, CA 94025

[GitHub](https://github.com/langgenius/dify) | [LinkedIn](https://www.linkedin.com/company/langgenius/posts/?feedView=all) | [X](https://x.com/dify_ai) | [Discord](https://discord.com/invite/FngNHpbcY7) | [Docs](https://docs.dify.ai/en/use-dify/getting-started/introduction)

You’re receiving this email because you signed up for Dify Cloud | [Unsubscribe](https://gdgfbbb.r.af.d.sendibt2.com/tr/un/v2/9vRNpGCjwSMmP8ZCc9xNjVwDUn6IA4feDgvXVRoFH_Hfv3X0U7NBWkKYlAwVVpVCsQEvEWdMQvKwfmmM3QlNtnjm38JCsL_OdEkCL9uGmfAiBOnb2jdCnaEwtOxdowRYLbOBdSmn-TIHDvzeXulkRtMKmXAmxF0HOMyLwMBCHd1rU1U_znjkJTKAsBWCjaQsw8T8jhakZc8fyrZ_4kX8lN7sYnuAP3Jbkmkzwfg2Z0AQa_nKoVUcg1sSEn6HROvSn-BUgqfmZXW69i8tBekQD_-nR-JqsFhxExufef9Me_RWgBw74B-37A7albY1aib4shqpirGCkonkQUJWIS5h-RFVGq10GCfC)

Accession

nse/dify/2026-10-20/4601

From

Anne from Dify · welcome@dify.ai

Sent

20 October 2026 · 21:00 UTC