OpenAI pipe returns kind of test data rather then real response data in pipe response

App Development Data Builder
1

Hi,

I am using openai api model “model”: “gpt-4o-2024-08-06” to return structured resume in response based on raw resume data in Sandbox environment.

When I use, the openai pipe test, It gives correct relevant data as output.
But when I use pipe response, and map it to table column as string in table rule, It has structured but kind of test data with the name of John Doe or Jane Doe.
Please find below the kind of response I am receiving as following:

“content”: {“personal_info”:{“name”:“Jane Doe”,“address”:“123 Maple Street, Springfield, IL 62704",“phone”:“555-123-4567”,“email”:"jane.doe@example.com”,“linkedin”:“linkedin.com/in/janedoe"},“professional_experience”:[{“company”:"Tech Solutions Inc.”,“position”:“Software Developer”,“duration”:“June 2018 - Present”},{“company”:“Innovatech Corp.”,“position”:“Junior Developer”,“duration”:“January 2016 - May 2018”}],“technical_skills”:[{“technology”:“Java”,“skill_name”:“Programming”},{“technology”:“SQL”,“skill_name”:“Database Management”},{“technology”:“JavaScript”,“skill_name”:“Web Development”}],“certifications”:[{“certification_name”:“Certified Java Developer”,“issuing_organisation”:“Oracle”,“issue_date”:“March 2019”}],“languages”:[{“language_name”:“English”,“proficiency_level”:“Native”},{“language_name”:“Spanish”,“proficiency_level”:“Intermediate”}],“projects”:[{“project_name”:“E-commerce Platform”,“description”:“Developed a full-featured e-commerce platform using Java and Spring Boot.”,“technologies_used”:“Java, Spring Boot, SQL”}],“education”:[{“degree”:“Bachelor of Science in Computer Science”,“institution”:“University of Illinois”,“graduation_year”:“2015”}],“tags”:[“Java”,“Software Development”,“Web Development”,“SQL”,“Project Management”],“extraction_percentage”:95.0}

Earlier I had used couple of files with the name of John Doe. But now with other names and different raw dataset, it is giving output with John Doe or Jane Doe with slightly different data everytime.

What could be the reason for this? Is it because I am using Tadabase sandbox environment or pipe is returning some kind of placeholder data because of any issue or Tadabase has some kind of chaching?

Awaiting response.

Thanks,
Khalid

2

I think the problem might be the response time of OpenAI, so you would need to add a timer, so OpenAI has enough time to generate the response, before you use the response to generate additions to the data tables.

I hope that solves the problem.