Azure AI - NLP Implementation
Budget: $30 – $250 USD
I am looking for an expert in Microsoft Azure, particularly in employing its AI capabilities for data analysis using Natural Language Processing (NLP). The project will involve implementing various NLP features such as text classification, sentiment analysis, and text-to-speech conversion.
Key Requirements:
- Proven experience with Microsoft Azure's AI and NLP capabilities.
- Ability to implement and customize NLP features according to project needs.
- Strong understanding of AI-driven data analysis.
Ideal Skills:
- Proficiency in text classification techniques.
- Expertise in sentiment analysis.
- Experience with text-to-speech conversion implementation.
- Prior experience with Microsoft Azure is a must.
I know this is very simple to set up but for the life of me i keep getting tripped up in the http or dynamic part i have tired to do this in power automate and Microsoft azzure and Ai foundry i have used open Ai deep research step by step and also openai operator however to no avail i almost want to quit and Jsut move the files to google driive as they can do it with Jsut a prompt but i put so much time learning i dont know code and have been tryjng to learn via youtube and the slow apoosch I understand that setting up this is quite straightforward, but I keep getting confused with the HTTP or dynamic part. I’ve tried doing this in Power Automate, Microsoft Azure, and AI Foundry. I’ve also used OpenAI’s Deep Research step-by-step guide and the OpenAI Operator, but nothing seems to work. I’m almost at the point of giving up and just moving the files to Google Drive, as they can do it with just a prompt. However, I’ve invested a lot of time learning code, and I don’t want to give up now. I’ve been trying to learn through YouTube videos and the slow approach, but it’s not working out.
Please assist me. Below is a detailed and visually organized outline that illustrates how your completed Logic App should appear once finalized. This reference will help ensure that each element is correctly positioned in the workflow:
⸻
✅ **Completed Logic App Visual Flow**
The structure of your Logic App designer (workflow) will clearly display these steps arranged from top to bottom:
? **Trigger:**
- **Event**: When a file is created in OneDrive
- **Folder**: IncomingDocs
⬇️
? **Action #1:**
- **Function**: Analyze Document using Azure AI Document Intelligence (Form Recognizer)
- **Parameters**:
- **Model Identifier**: prebuilt-read
- **File Content**: Utilize the dynamic content (File content) sourced from the previous OneDrive trigger step.
⬇️
? **Action #2 (HTTP Call):**
- **Purpose**: Generate an Employee Statement using GPT-4
- **Request Type**: HTTP POST directed to Azure OpenAI endpoint
- **Payload**: JSON Body that includes dynamic text obtained from the Document Intelligence analysis.
⬇️
? **Action #3 (OneDrive Save):**
- **Operation**: Save the Employee Statement to OneDrive
- **Target Folder**: Reports
- **File Naming**: EmployeeStatement.docx (or .txt)
- **File Content**: Ingest dynamic content generated from the GPT-4 action above.
⬇️
? **Action #4 (HTTP Call):**
- **Purpose**: Generate a Medical Report using GPT-4
- **Request Type**: HTTP POST to Azure OpenAI endpoint
- **Payload**: JSON Body that integrates dynamic text from the earlier Document Intelligence step.
⬇️
? **Action #5 (OneDrive Save):**
- **Operation**: Save the Medical Report to OneDrive
- **Target Folder**: Reports
- **File Naming**: MedicalReport.docx (or .txt)
- **File Content**: Utilize the dynamic content from the GPT-4 step listed above.
⬇️
? **Action #6 (HTTP Call):**
- **Purpose**: Generate the Final Claim Report using GPT-4
- **Request Type**: HTTP POST to Azure OpenAI endpoint
- **Payload**: JSON Body incorporating dynamic text from Document Intelligence.
⬇️
? **Action #7 (OneDrive Save):**
- **Operation**: Save the Final Claim Report to OneDrive
- **Target Folder**: Reports
- **File Naming**: FinalClaimReport.docx (or .txt)
- **File Content**: Draw from the dynamic content produced in the previous GPT-4 action.
⬇️
? **Action #8 (Move Original File):**
- **Function**: Archive the original file
- **Target Folder**: Archive
⬇️
? **Action #9 (Email Notification - Optional):**
- **Action**: Send an Email (Outlook)
- **Condition**: Triggers only if an error occurs, notifying of the failure.
⸻
? **Visual Workflow Overview**:
[Trigger: File Created in OneDrive]
│
└─► [Analyze Document with Intelligence]
│
├─► [Generate Employee Statement (GPT-4)] ──► [Save Employee Statement to OneDrive]
│
├─► [Generate Medical Report (GPT-4)] ──────► [Save Medical Report to OneDrive]
│
└─► [Generate Final Claim Report (GPT-4)] ──► [Save Final Claim Report to OneDrive]
│
└─► [Move Original File to Archive]
│
└─► [Send Email Notification (On Error Only)]
⸻
? **Visual Confirmation Checklist**:
- Each step is clearly labeled for easy navigation.
- All HTTP steps are directly linked to the Azure OpenAI service.
- Following each GPT-4 action, there's a corresponding step for saving files.
- Each file-saving procedure explicitly shows that it is directed to the “Reports” folder.
- The archiving action methodically indicates the movement of the original file.
- The optional email notification is distinctly marked and is set to activate only in the case of failures.
⸻
✅ **Next Steps for Visual Confirmation**:
- After saving your Logic App, navigate to the Run history to visually confirm that each step executed successfully, marked by green check marks ✅.
- If any issues arise, the respective step will display a red exclamation mark ❌, allowing you to click for debugging purposes.
This comprehensive visual guide ensures that your Logic App workflow aligns perfectly with your intended objectives and facilitates easy step-by-step verification. I’ve been trying to implement this myself, but I am facing challenges in Microsoft, and I really need a developer’s assistance in order to successfully carry out this task.
Key Requirements:
- Proven experience with Microsoft Azure's AI and NLP capabilities.
- Ability to implement and customize NLP features according to project needs.
- Strong understanding of AI-driven data analysis.
Ideal Skills:
- Proficiency in text classification techniques.
- Expertise in sentiment analysis.
- Experience with text-to-speech conversion implementation.
- Prior experience with Microsoft Azure is a must.
I know this is very simple to set up but for the life of me i keep getting tripped up in the http or dynamic part i have tired to do this in power automate and Microsoft azzure and Ai foundry i have used open Ai deep research step by step and also openai operator however to no avail i almost want to quit and Jsut move the files to google driive as they can do it with Jsut a prompt but i put so much time learning i dont know code and have been tryjng to learn via youtube and the slow apoosch I understand that setting up this is quite straightforward, but I keep getting confused with the HTTP or dynamic part. I’ve tried doing this in Power Automate, Microsoft Azure, and AI Foundry. I’ve also used OpenAI’s Deep Research step-by-step guide and the OpenAI Operator, but nothing seems to work. I’m almost at the point of giving up and just moving the files to Google Drive, as they can do it with just a prompt. However, I’ve invested a lot of time learning code, and I don’t want to give up now. I’ve been trying to learn through YouTube videos and the slow approach, but it’s not working out.
Please assist me. Below is a detailed and visually organized outline that illustrates how your completed Logic App should appear once finalized. This reference will help ensure that each element is correctly positioned in the workflow:
⸻
✅ **Completed Logic App Visual Flow**
The structure of your Logic App designer (workflow) will clearly display these steps arranged from top to bottom:
? **Trigger:**
- **Event**: When a file is created in OneDrive
- **Folder**: IncomingDocs
⬇️
? **Action #1:**
- **Function**: Analyze Document using Azure AI Document Intelligence (Form Recognizer)
- **Parameters**:
- **Model Identifier**: prebuilt-read
- **File Content**: Utilize the dynamic content (File content) sourced from the previous OneDrive trigger step.
⬇️
? **Action #2 (HTTP Call):**
- **Purpose**: Generate an Employee Statement using GPT-4
- **Request Type**: HTTP POST directed to Azure OpenAI endpoint
- **Payload**: JSON Body that includes dynamic text obtained from the Document Intelligence analysis.
⬇️
? **Action #3 (OneDrive Save):**
- **Operation**: Save the Employee Statement to OneDrive
- **Target Folder**: Reports
- **File Naming**: EmployeeStatement.docx (or .txt)
- **File Content**: Ingest dynamic content generated from the GPT-4 action above.
⬇️
? **Action #4 (HTTP Call):**
- **Purpose**: Generate a Medical Report using GPT-4
- **Request Type**: HTTP POST to Azure OpenAI endpoint
- **Payload**: JSON Body that integrates dynamic text from the earlier Document Intelligence step.
⬇️
? **Action #5 (OneDrive Save):**
- **Operation**: Save the Medical Report to OneDrive
- **Target Folder**: Reports
- **File Naming**: MedicalReport.docx (or .txt)
- **File Content**: Utilize the dynamic content from the GPT-4 step listed above.
⬇️
? **Action #6 (HTTP Call):**
- **Purpose**: Generate the Final Claim Report using GPT-4
- **Request Type**: HTTP POST to Azure OpenAI endpoint
- **Payload**: JSON Body incorporating dynamic text from Document Intelligence.
⬇️
? **Action #7 (OneDrive Save):**
- **Operation**: Save the Final Claim Report to OneDrive
- **Target Folder**: Reports
- **File Naming**: FinalClaimReport.docx (or .txt)
- **File Content**: Draw from the dynamic content produced in the previous GPT-4 action.
⬇️
? **Action #8 (Move Original File):**
- **Function**: Archive the original file
- **Target Folder**: Archive
⬇️
? **Action #9 (Email Notification - Optional):**
- **Action**: Send an Email (Outlook)
- **Condition**: Triggers only if an error occurs, notifying of the failure.
⸻
? **Visual Workflow Overview**:
[Trigger: File Created in OneDrive]
│
└─► [Analyze Document with Intelligence]
│
├─► [Generate Employee Statement (GPT-4)] ──► [Save Employee Statement to OneDrive]
│
├─► [Generate Medical Report (GPT-4)] ──────► [Save Medical Report to OneDrive]
│
└─► [Generate Final Claim Report (GPT-4)] ──► [Save Final Claim Report to OneDrive]
│
└─► [Move Original File to Archive]
│
└─► [Send Email Notification (On Error Only)]
⸻
? **Visual Confirmation Checklist**:
- Each step is clearly labeled for easy navigation.
- All HTTP steps are directly linked to the Azure OpenAI service.
- Following each GPT-4 action, there's a corresponding step for saving files.
- Each file-saving procedure explicitly shows that it is directed to the “Reports” folder.
- The archiving action methodically indicates the movement of the original file.
- The optional email notification is distinctly marked and is set to activate only in the case of failures.
⸻
✅ **Next Steps for Visual Confirmation**:
- After saving your Logic App, navigate to the Run history to visually confirm that each step executed successfully, marked by green check marks ✅.
- If any issues arise, the respective step will display a red exclamation mark ❌, allowing you to click for debugging purposes.
This comprehensive visual guide ensures that your Logic App workflow aligns perfectly with your intended objectives and facilitates easy step-by-step verification. I’ve been trying to implement this myself, but I am facing challenges in Microsoft, and I really need a developer’s assistance in order to successfully carry out this task.