Resume Parser (JSON Extractor)

A high-precision resume parsing model built on top of LLaMA 3.1 using Ollama. This model extracts structured information from raw resume text and returns clean, schema-based JSON.

🚀 What it does

• Converts unstructured resume text → structured JSON

• Extracts key sections like:

  • Personal Information
  • Work Experience
  • Education
  • Skills
  • Projects & Certifications

• Ensures consistent and machine-readable output

⚙️ Usage

ollama run yourusername/resume-parser

Paste the resume text when prompted.

📥 Input

Provide raw resume text (from PDF, DOCX, etc.)

Example:

John Doe
Email: john@example.com
Phone: +91-9876543210

Experience:
Software Engineer at XYZ (2022–Present)
- Built APIs using FastAPI

📤 Output

The model returns ONLY valid JSON in the following format:

 {
    "_id": "1627b8c0-5466-4393-b015-0bbe579b11f7",
    "filename": "resume-sample.pdf",
    "extracted_at": "2026-03-16T14:36:23.426188",
    "page_count": 13,
    "char_count": 32663,
    "data": {
      "personal_info": {
        "full_name": "IM A. SAMPLE I",
        "email": "imasample1@xxx.com",
        "phone": "+1 402 292-2345",
        "location": {
          "city": "Bellevue",
          "state": "Nebraska",
          "country": "United States",
          "full_address": "1234 North 55 Street, Bellevue, Nebraska 68005"
        },
        "linkedin_url": null,
        "github_url": null,
        "portfolio_url": null,
        "other_urls": []
      },
      "professional_summary": null,
      "objective": "Internship or Part-time Position in Marketing, Public Relations or related field utilizing strong academic background and excellent communication skills",
      "work_experience": [
        {
          "job_title": "Student Intern, Financial Accounting Development Program",
          "company_name": "Mutual of Omaha",
          "location": "Omaha, NE",
          "employment_type": "Internship",
          "start_date": "Summer 20xx",
          "end_date": null,
          "is_current": false,
          "duration_months": null,
          "responsibilities": [
            "Developed and maintained accounting records for up to fifty bank accounts.",
            "Formulated monthly and year-end financial statements and generated various payroll records, including federal and state payroll reports, annual tax reports, W-2 and 1099 forms, etc.",
            "Tested accuracy of account balances and prepared supporting documentation for submission during a comprehensive three-year audit of financial operations.",
            "Formulated intricate pro-forma budgets.",
            "Calculated and implemented depreciation/amortization schedules."
          ],
          "achievements": [
            "Converted manual to computerized accounting systems for two organizations.",
            "Analyzed and successfully reprogrammed software to meet customer requirements.",
            "Researched and corrected problems to assure effective operation of newly computerized systems."
          ],
          "technologies_used": [
            "MS Office (Word, Excel, PowerPoint, Outlook), QuickBooks"
          ]
        }

⚠️ Rules

• Output is STRICT JSON only
• No explanations or extra text
• Missing fields → null or []
• No hallucinated data

🧠 Model Info

• Base Model: llama3.1:8b
• Type: Prompt-engineered (no fine-tuning)
• Runtime: Ollama (local)

💡 Use Cases

• Resume screening systems (ATS)
• HR automation tools
• Candidate data extraction pipelines
• AI-powered recruitment platforms

⚡ Tips

• Use clean resume text for best results
• Set temperature = 0 for consistent output
• Validate JSON before storing

📌 Note

This model uses prompt engineering for structured extraction. Accuracy depends on input quality and resume format.

⭐ Support

If you find this useful, consider giving it a ⭐ and sharing it!