Skip to main content

Job Profile Parser

Table of Contents​

  1. Overview
  2. Getting Started
  3. Input Data Requirements
  4. How to Prompt the Agent
  5. Example Usage
  6. Best Practices
  7. Troubleshooting
  8. FAQ

Overview​

The Job Profile Data Parser Agent is a specialized extraction and parsing tool designed to process uploaded documents containing job profile information. This agent excels at identifying, extracting, and structuring job profile data from various document formats into standardized CSV outputs for further processing and analysis.

Key Capabilities​

  • Extracts multiple job profiles from single documents
  • Parses structured and unstructured job profile information
  • Converts complex document formats to standardized CSV
  • Preserves hierarchical task structures and multi-line content
  • Maintains data integrity without inferring missing information
  • Supports batch processing of multiple profiles
  • Provides visual data preview through dataframes

Core Extraction Fields​

  • Job title and code
  • Description and department
  • Category and subcategory classifications
  • Industry applications
  • Proficiency levels
  • Prerequisites and skills
  • Task structures (including tables)

Getting Started​

Prerequisites​

  • Structured document containing job profile information

Input Data Requirements​

Supported Document Types​

  1. Structured Documents

    • PDF files with job profiles
    • Word documents with formatted content
    • Excel sheets with job data
    • HTML documents with profile information

  2. Content Formats

    • Formal job descriptions
    • Framework documentation
    • Competency models
    • Task analysis documents
    • Skill inventories

Data Structure Requirements​

Documents should contain:

  • Clear job profile identifiers
  • Distinguishable sections or fields
  • Consistent formatting (preferred)
  • Complete profile information

Field Specifications​

  • Title: Job profile name/designation
  • Code: Unique identifier or reference number
  • Description: Role overview (brief)
  • Department: Organizational unit
  • Category: Primary classification
  • Subcategory: Secondary classification
  • Industry: Applicable sectors
  • Level: Proficiency/seniority level
  • Prerequisites: Required qualifications
  • Skills: Associated competencies
  • Tasks: Responsibilities (including tables)

How to Prompt the Agent​

Effective Prompt Structure​

Basic Extraction​

"Extract job profiles from the uploaded document"

Detailed Extraction Request​

"Parse all job profiles from the attached HR framework document. 
Ensure you capture:
- All task tables with complete column data
- Multi-level skill hierarchies
- Industry-specific variations"

Multiple Profile Processing​

"Extract and structure all job profiles found in the document.
Maintain the original formatting for tasks and preserve all subcategories."

Specific Field Focus​

"Parse job profiles with emphasis on:
- Complete task descriptions including tabular data
- All prerequisite requirements
- Skill categorizations
Focus on maintaining data structure integrity."

Prompt Best Practices​

  1. Upload document before requesting extraction
  2. Specify if certain fields are priority
  3. Mention if table structures need preservation
  4. Indicate if multiple profiles are expected
  5. Request specific handling for complex structures

Example Usage​

Example 1: Single Profile Extraction​

Input Document: PDF with one detailed job profile Prompt:

"Extract the job profile from the uploaded PDF document"

Expected Output:

  • CSV with single row containing all fields
  • Dataframe preview showing structured data
  • Downloadable CSV file
  • Confirmation of successful extraction

Example 2: Framework Document Processing​

Input Document: Competency framework with 10 job profiles Prompt:

"Parse all job profiles from the competency framework document. 
Capture complete task tables and skill mappings."

Expected Output:

  • CSV with 10 rows (one per profile)
  • Multi-line content preserved in cells
  • Complete task structures maintained
  • Visual dataframe display
  • Export file with all profiles

Example 3: Complex Task Table Extraction​

Input Document: Document with job profiles containing detailed task tables Prompt:

"Extract job profiles ensuring all task table rows and columns are captured completely"

Expected Output:

  • Tasks field containing full table structure
  • Line breaks preserved with \n
  • All columns represented
  • Structured CSV output

Best Practices​

1. Document Preparation​

  • Ensure documents are readable and not corrupted
  • Use high-quality scans for PDF documents
  • Maintain consistent formatting where possible
  • Include clear section headers

2. Extraction Optimization​

  • Upload one document at a time for clarity
  • Specify expected number of profiles
  • Mention any unique formatting requirements
  • Request specific field priorities if needed

3. Data Validation​

  • Review dataframe preview before export
  • Check for missing critical fields
  • Verify multi-line content preservation
  • Confirm profile count matches expectations

4. Output Handling​

  • Download CSV immediately after generation
  • Verify delimiter usage (; by default)
  • Check encoding for special characters
  • Validate against source document

5. Complex Structures​

  • Explicitly mention table preservation needs
  • Specify hierarchy maintenance requirements
  • Request line break preservation
  • Indicate multi-value field handling

Troubleshooting​

Common Issues and Solutions​

Issue: Missing Fields in Output​

Symptom: Some expected fields are empty Solution:

  • Agent only extracts explicitly present data
  • Check if fields exist in source document
  • Fields left empty if not found (no inference)

Issue: Table Structure Lost​

Symptom: Task tables appear as single line Solution:

  • Explicitly request table structure preservation
  • Ensure source document has clear table formatting
  • Check for \n line breaks in output

Issue: Multiple Profiles Not Detected​

Symptom: Only first profile extracted Solution:

  • Ensure profiles are clearly delineated in document
  • Request "all profiles" explicitly
  • Check document structure for consistency

Issue: CSV Format Issues​

Symptom: Data not properly delimited Solution:

  • Default delimiter is ;
  • Ensure no delimiter conflicts in data
  • Values are enclosed in double quotes

Issue: Special Characters Corrupted​

Symptom: Encoding issues with special characters Solution:

  • Check file encoding (UTF-8 recommended)
  • Verify source document character encoding
  • Re-export with proper encoding specified

Issue: Empty Extraction Results​

Symptom: No data extracted from document Solution:

  • Verify document upload successful
  • Check document readability
  • Ensure document contains job profile data
  • Try different document format

FAQ​

Q: Does the agent infer missing information?​

A: No, the agent strictly extracts only explicitly present data. Missing fields are left empty rather than inferred.

Q: How are multi-line fields handled?​

A: Multi-line content is preserved using \n line breaks within quoted CSV cells.

Q: Can it process multiple documents simultaneously?​

A: Process one document at a time for best results and clarity.

Q: What's the maximum number of profiles per document?​

A: No hard limit, but performance is optimal with up to 50 profiles per document.

Q: How are task tables preserved?​

A: Complete table structures are captured with all rows and columns, maintaining relationships through formatting.

Q: What happens to hierarchical data?​

A: Hierarchies are flattened but relationships are preserved through formatting and line breaks.

Q: Can it extract from images?​

A: No, documents must contain machine-readable text. OCR preprocessing may be needed for scanned images.

Q: How does it handle duplicate profiles?​

A: Each profile instance is extracted as a separate row, even if duplicated.

Q: What about non-standard field names?​

A: The agent maps variations to standard fields where possible, but unusual fields may not be captured.

Q: Can I customize the output fields?​

A: The field structure is standardized, but you can request emphasis on specific fields during extraction.

Q: How accurate is the extraction?​

A: Accuracy depends on document quality and structure. Well-formatted documents yield near-perfect extraction.

Q: What file formats can be uploaded?​

A: PDF, Word, Excel, HTML, and text files containing job profile information.

Q: Is there validation for extracted data?​

A: Basic structure validation is performed, but content accuracy should be verified against source.

Q: Can it merge data from multiple sources?​

A: No, each document is processed independently. Merging should be done post-extraction.