Skip to content

Project Work & WASP1 Grading Letters

Overview

This tool generates formal grading letters for project work (Praxisprojekt), WASP1 projects, and similar academic work at TH Köln. It extracts metadata from the project PDF, automatically determines the appropriate formal address, and generates a LaTeX letter template with TH Köln formatting.

Generated Files (Results)

A detailed overview with images can be found on the Examples page.

1. LaTeX Grading Letter

Filename: projektarbeit_brief_<matrikelnr>.tex A formal scrlttr2 letter with TH Köln letterhead and footer.

2. Compiled PDF

Filename: projektarbeit_brief_<matrikelnr>.pdf The ready-to-print PDF of the grading letter (requires LuaLaTeX).

3. Email for Examination Office

Filename: projekt_anmeldung_<name>_<matrikelnr>.md A ready-to-use email template for submitting the grade to the examination office.

4. Student Feedback Email

Filename: feedback_student_<name>_<matrikelnr>.md An automatically generated draft containing a summary of strengths/weaknesses and the final grade.

5. Web Metadata (Profile)

Filename: YYYY-MM-DD-title.md A Jekyll-compatible profile of the work for your own website. The path can be defined globally in config.yaml.

Requirements

  • Project work PDF with a title page (selectable text required)
  • At least one configured LLM API key (OpenAI, Groq, Google Gemini) or local Ollama
  • LaTeX installation (LuaLaTeX recommended)
  • (Optional) Signature image at data/signature.png

Usage

The recommended way is to use a configuration file.

Command Line (CLI)

# Using a config file
academic-doc-generator --config config_project_template.json

# Basic usage
academic-doc-generator project /path/to/Praxisprojekt_Mueller.pdf

Usage via main.py

You can also run the tool via main.py by specifying the path to the project folder there. Details can be found in the Configuration Guide.

Feedback & Student Email

When create_feedback_mail is enabled (default), the tool: 1. Analyzes the project work using the LLM. 2. Extracts the student's email address from the title page. 3. Generates a Markdown file with a feedback summary. 4. Provides a draft email for the student.

How It Works (Details)

  1. Metadata Extraction: Automatically detects student name, ID, title, examiner, and student email.
  2. Gender Detection: Uses an LLM to determine the appropriate formal German address (Herr/Frau) based on the first name.
  3. Semester Calculation: Automatically determines the current semester (SoSe/WS).
  4. LaTeX Generation: Creates a professional letter with TH Köln formatting.
  5. Signature Integration: Automatically includes your signature image if available.

Troubleshooting

Student Email Not Extracted

  • Ensure the email address is present on the title page and is selectable text.
  • If it fails, you will need to add the recipient manually to the email draft.

Incorrect Salutation (Mr/Ms)

  • For ambiguous names, the detection might be incorrect. You can easily fix the salutation in the generated .tex file.