How tonykipkemboi/resume-optimization-crew works
Main Components
• Input and Configuration:
• Environment Setup: The .env file holds API keys.
• Resume Input: A PDF resume is stored in the “knowledge” directory.
• Job and Company Data: Specified as parameters inside main.py (job URL and company name).
• Core Application (src/resume_crew):
• Main Entry Point: main.py – Initiates the optimization process.
• Configuration Files: YAML files (agents.yaml, tasks.yaml) in the “config” folder provide settings for the AI agents.
• AI Agents Modules:
▪ Job Analyzer – Responsible for job analysis (extracts requirements and computes match scores).
▪ Resume Analyzer – Analyzes and scores the resume, then suggests improvements.
▪ Company Researcher – Conducts research on the company for interview preparation.
• Supporting Tools: Custom tool(s) provided in the “tools” directory.
• Output Section:
• Output Directory: Contains JSON and Markdown files that hold job analysis, optimized resume suggestions, and company research data.
Relationships and Interactions
• Start at the Inputs:
– The process begins with the resume PDF and the input manifest (job URL and company name) defined in main.py.
– Environment variables from .env include API keys needed for the AI interactions.
• Main Flow (Central Processing):
– The main script (main.py) orchestrates the execution: it reads configurations from the YAML files and then dispatches data to the three AI agents.
1. Job Analyzer (Extracts and analyzes job details)
2. Resume Analyzer (Calculates match scores and recommends improvements)
3. Company Researcher (Gathers company information)
• Aggregation and Output:
– Once the agents perform their tasks, their results are collected.
– The results are then saved as multiple files in the “output” folder (e.g., job_analysis.json, resume_optimization.json, company_research.json) and final reports such as final_report.md.
Architectural Patterns and Technologies
• The project adopts a modular design where a central script (main.py) orchestrates multiple AI agents. This pattern resembles a microservice style within a single code base, where each agent performs a specialized task.
• External dependencies (OpenAI API, Serper API) that are injected via environment variables.
• The CrewAI tool is central to the project, so ensure its role is clear in your flowchart when showing the agent-based operations.
• Every component is clearly labeled with its role and corresponding file/folder as per the file structure.
– From input (resume PDF, job URL, API keys) to main.py.
– From main.py outward to the three distinct agents.
– From each agent to outputs in “output/”.
A Sample Flowchart Layout (Text-Based Representation):
[Input: Resume PDF, Job URL, API Keys]
│
▼
[Main Controller (main.py)]
│
┌┴─────────┬─────────┐
▼ ▼ ▼
[Job Analyzer] [Resume Analyzer] [Company Researcher]
(AI Agent) (AI Agent) (AI Agent)
│ │ │
└───┬──────┴───────┬───┘
▼ ▼
[Aggregate & Process Results]
│
▼
[Output: JSON/Markdown Files]
(output/ directory)
This flowchart clearly shows the progression from inputs through central orchestration to specialized processing and finally to the outputs—all of which capture the project’s modular, AI-agent driven architecture.