How pranavbp525/resumegenerator works
Main Components
• Streamlit Frontend (app.py): This is the entry point for users. It presents a user interface that accepts job descriptions and triggers resume generation.
• Resume Generation Core (modules in src/):
– llm.py: Manages interactions with the OpenAI API.
– prompts.py: Defines templates and prompt setups for generating resume content.
– responses.py: Handles parsing and processing of responses from the API.
• Template & Resume Configuration:
– assets/resume.yaml: Contains user-specific resume details that merge with generated data.
– LaTeX Template Files (within artifacts/*/latex_src): These are used to compile the final resume document.
• Artifact Management:
– artifacts/ directory: Stores generated outputs including intermediate LaTeX files and the final resume.pdf, along with model outputs in JSON.
• Additional Utilities and Experimentation:
– generate.py, experiments.ipynb, tree.py, and utils.py scripts: Likely support testing, additional generating flows, and system utilities.
Relationships and Interactions
• The UI layer (Streamlit app) gathers input (job description) from the user.
• The input is passed to the resume generation core, where prompts are crafted and sent to the OpenAI API (via llm.py), receiving generated content.
• The generated data is combined with the static resume details from assets/resume.yaml.
• The combined data flows into a LaTeX processing pipeline (using the LaTeX template files in the artifacts/*/latex_src directory) to produce a PDF resume.
• Outputs are saved in the artifacts folder, and possibly feedback or logging details are generated.
Architectural Patterns and Principles
• Layered Architecture: Separation of concerns between the user interface (Streamlit), business logic (resume generation in the src modules), and storage (assets for configurations; artifacts for outputs).
• Modularity: Distinct modules for prompts, API communication, and response handling.
• External Service Integration: The use of OpenAI’s API is a key dependency and is scoped within the llm module.
Technologies, Frameworks, and Libraries
• Python with Streamlit for the front end.
• Pipenv for dependency management.
• OpenAI API for language model-based generation.
• LaTeX for resume formatting and PDF generation.
• YAML for configuration (assets/resume.yaml).
Elements
– “User Interface (Streamlit App – app.py)”
– “Resume Generation Core” (with subcomponents for API integration, prompt handling, and response processing from src/llm.py, src/prompts.py, and src/responses.py)
– “Configuration & Templates” (assets/resume.yaml and LaTeX template files under artifacts/*/latex_src)
– “Artifact Storage” (final resume.pdf and intermediate model_outputs.json)
– Flow of job description from UI to the generation core.
– Flow from the generation core to OpenAI API and back.
– Flow of generated data merging with YAML configuration into the LaTeX template for final PDF generation.