Resume Parser API

---

THIS IS NOT THE MATCHING ENGINE

---

Steps to Run

1. Install all dependencies from requirements.txt.
2. Activate the virtual environment (venv).
3. Start the API.

---

Current API Routes

1. (POST) /resume/pdf

   • Description: Upload your resume as form-data.
     • Key: file
     • MIME Type: application/pdf
   • Output:

     {
         "filename": "da.pdf",
         "status": "Received file and saved successfully"
     }

2. (GET) /predict/<filename.pdf>

   • Description: Fetch the predicted output for the uploaded file.
   • Output:

     {
         "entities": [
             {
                 "label": "Entity 1",
                 "text": "Entity Value"
             },
             {
                 "label": "Entity 2",
                 "text": "Entity Value"
             }
         ]
     }

3. (NEW!!!) (POST) /resume/beta

   • Description: Newer, Optimised and Faster Method to get predictions.
     • Key: file
     • MIME Type: application/pdf
   • Output:

     {
         "name": " ",
         "email": "",
         "mobile_number": "",
         "skills": [],
         "college_name": null,
         "degree": [""],
         "designation": [""],
         "experience": [""],
         "company_name": null,
         "no_of_pages": 2,
         "total_exp": null,
         "total_experience": 0.0,
         "linkedin": null
     }

4. (POST) /matching/short

   • Description: Processes the provided JSON data and returns profile and relevancy scores.
   • Output:

     {
         "profile_score": 85.0,
         "relevancy_score": 90.0
     }

5. (POST) /matching/longVerbose

   • Description: Processes the provided JSON data and returns detailed profile comparison information.
   • Output:

     {
         "profile_score": 85.0,
         "relevancy_score": 90.0,
         "candidates": [
             {
                 "name": "John Doe",
                 "intersection_score": 75.0,
                 "cosine_similarity": 85.0,
                 "jaccard_similarity": 60.0,
                 "overall_similarity": 70.0
             },
             {
                 "name": "Jane Doe",
                 "intersection_score": 65.0,
                 "cosine_similarity": 75.0,
                 "jaccard_similarity": 55.0,
                 "overall_similarity": 60.0
             }
         ]
     }

---

Notes

• The current implementation uses V1 of the model.
• A new model is under development, but the output structure will remain the same.
• Enjoy using the current version in the meantime!

---

To-Do List

• Add OCR support using Tesseract.
• Add Matching Engine.
• Add JD Parser.
• Your Mom .
  \