Not able to get flow to run - Gummie ran me around for two hours. Please help

[Garbanzo]

Here is my use case.
I have 150 small png files that are screenshots of contact data from a conference app.
I wanted to get a CSV table of all the contacts that are contained in the screenshots. There are a lot of screenshots. I basically am trying to scrap the conference app when they don’t let you download the data but you can scroll through the app and screenshot it.

I designed the flow as follows

1. Google drive reader
2. Analyze image
3. ask ai to format data
4. write to csv or google sheet.

If you look at the Gummie help history you see we went in circles for about tow hours.
The problem is the output from the Ask AI was a single out put and the input expect was for the 4 columns of contact data.

1. first name
2. last name
3. company
4. title when available

https://www.gumloop.com/pipeline?workbook_id=pMcs35ANzG1c6B7mYjwWuw&run_id=KkVxTGCZriJ2f2C6jRaYmt

[Gumloop_Bot]

Hey @Garbanzo! If you’re reporting an issue with a flow or an error in a run, please include the run link and make sure it’s shareable so we can take a look.

1. Find your run link on the history page. Format: https://www.gumloop.com/pipeline?run_id={your_run_id}&workbook_id={workbook_id}

2. Make it shareable by clicking “Share” → ‘Anyone with the link can view’ in the top-left corner of the flow screen.
   

3. Provide details about the issue—more context helps us troubleshoot faster.

You can find your run history here: https://www.gumloop.com/history

[Shrikar]

Hey @Garbanzo is that a web app or mobile app? Do you mind sharing that content? Also from the error it was able to extract the content but writing to notion failed. Only one thing you could do is use Extract data note to convert from text to 4 different columns you need and then write to notion.

[Wasay-Gumloop]

Hey @Garbanzo - Shrikar’s response above is spot on! You can pass the AI response from the Analyze Image node to an Extract Data node to get separate outputs for each field. Here’s an example: https://www.gumloop.com/pipeline?workbook_id=8rgxNQ9p8aHznSj2X1nmmA

https://www.gumloop.com/pipeline?workbook_id=pMcs35ANzG1c6B7mYjwWuw&run_id=KkVxTGCZriJ2f2C6jRaYmt

The error here is actually related to Notion’s API. You can wrap the node in an Error Shield to prevent this from halting the entire flow.

[Garbanzo]

Thanks for your help. I only put Notion in because I got frustrated. Gummie is giving me instructions that I can’t complete.
Here’s the step-by-step process:

1. First, add the Analyze Image node to your canvas:

   • Set the “Prompt” parameter to something like: “Extract all contact information from this image, including names, phone numbers, email addresses, and physical addresses. List all information clearly.”
   • You can either upload an image file or use an image link

2. Connect the Analyze Image node’s “Analysis” output to an Extract Data node:

   • Enable “Extract List?” if you have multiple contacts
   • Define data fields like:
     • Name (text)
     • Phone (text)
     • Email (text)
     • Address (text)
   • Add “Additional Context” explaining these are contact details from business cards or documents

3. Connect the Extract Data node’s outputs to an Extract to Table node:

   • Enable “Extract Multiple Rows?” if you have multiple contacts
   • Select or link to a Google Sheet
   • Make sure your sheet has headers matching your extracted data fields

This flow will analyze your image, extract the contact information in a structured format, and then place it into your Google Sheet table.

The problem is the output of the extract data node has 4 columns and the input of the extract to text only has one input.

This is the issue all along and Gummie had me keeping modifying to the same issue

Here is the shared pipeline.
https://www.gumloop.com/pipeline?workbook_id=29Xd9gjCRQ7vt9Td6UWxXB

[Wasay-Gumloop]

Hey @Garbanzo – You can use the Google Sheet Writer node directly here, here’s an example: https://www.gumloop.com/pipeline?workbook_id=ays4p35Ye2BGPuQqMYueWn

Let me know if that works for you.

Gummie’s response is also correct but it just requires one additional step which would be a Combine Text node to merge the data from all the data fields from the Extract Data node.