swarms/docs/examples/job_finding.md

AI Job Search Swarm

Overview

The AI Job Search Swarm is an intelligent multi-agent system designed to automate and streamline the job search process using the Swarms framework. It leverages specialized AI agents to analyze user requirements, execute comprehensive job searches, and curate relevant opportunities, transforming traditional job hunting into an intelligent, collaborative process.

Key Components

The Job Search Swarm consists of three specialized agents, each responsible for a critical stage of the job search process:

Agent Name	Role	Responsibilities
Sarah-Requirements-Analyzer	Extracts and clarifies job search requirements.	Engages with the user to understand job titles, skills, locations, salary, company preferences, experience level, work authorization, and career goals. Analyzes user responses to identify key search terms, must-haves, deal-breakers, and priority factors. Generates 3-5 optimized search queries.
David-Search-Executor	Executes job searches and analyzes results for relevance.	Uses the exa_search tool for each query provided. Analyzes results for job title, skills, location, salary, company reputation, responsibilities, and growth opportunities. Categorizes results by match strength (Strong, Good, Moderate, Weak) and extracts key job details.
Lisa-Results-Curator	Filters, organizes, and presents job search results.	Reviews all search results. Filters and prioritizes based on match scores, user requirements, and job quality. Organizes results into "Top Recommendations," "Strong Alternatives," and "Worth Considering." Provides detailed comparisons for top recommendations and engages the user for feedback to refine the search.

Installation

To run the AI Job Search Swarm, you'll need to set up your Python environment and configure necessary API keys.

Prerequisites

• Python 3.8+
• An Exa API Key

Installation Steps

1. Clone the repository (if you haven't already):

   git clone https://github.com/kyegomez/swarms.git
   cd swarms
   

2. Create a virtual environment and activate it:

   python -m venv venv
   source venv/bin/activate
   

3. Install dependencies: The job_finding.py script relies on several libraries. These can be installed using the requirements.txt file located at the root of the project.

   pip install -r requirements.txt
   

   This will install httpx, python-dotenv, loguru, and other necessary packages.

API Key Configuration

The exa_search tool used by the Search Executor Agent requires an EXA_API_KEY.

1. Obtain an Exa API Key: Visit Exa AI to get your API key.
2. Create a .env file: In the root directory of your project, create a file named .env and add your Exa API key to it:

   EXA_API_KEY="your_exa_api_key_here"
   

   Replace "your_exa_api_key_here" with your actual Exa API Key. The python-dotenv library will automatically load this key into your environment.

Example Usage

To run the job search swarm, execute the job_finding.py script from your terminal.

python examples/demos/apps/job_finding.py

The script is designed to run with a pre-filled user requirement for demonstration purposes. You can modify the user_requirements string in the main() function within examples/demos/apps/job_finding.py to test different job search scenarios.

Example User Requirements (from main() function):

from typing import List
import os
from dotenv import load_dotenv
from loguru import logger
import httpx
from swarms.structs.agent import Agent
from swarms.structs.conversation import Conversation
from swarms.utils.history_output_formatter import history_output_formatter
from swarms.utils.any_to_str import any_to_str

# --- Exa Search Tool Integration ---
def exa_search(
    query: str,
    characters: int = 500,
    sources: int = 3,
) -> str:
    """
    Perform a highly summarized Exa web search for job listings and career information.

    Args:
        query (str): Search query for jobs or career info.
        characters (int): Max characters for summary.
        sources (int): Number of sources.

    Returns:
        str: Condensed summary of search results.
    """
    api_key = os.getenv("EXA_API_KEY")
    if not api_key:
        raise ValueError("EXA_API_KEY environment variable is not set")

    headers = {
        "x-api-key": api_key,
        "content-type": "application/json",
    }

    payload = {
        "query": query,
        "type": "auto",
        "numResults": sources,
        "contents": {
            "text": True,
            "summary": {
                "schema": {
                    "type": "object",
                    "required": ["answer"],
                    "additionalProperties": False,
                    "properties": {
                        "answer": {
                            "type": "string",
                            "description": "Highly condensed summary of the search result",
                        }
                    },
                }
            },
            "context": {"maxCharacters": characters},
        },
    }

    try:
        logger.info(f"[SEARCH] Exa job search: {query[:50]}...")
        response = httpx.post(
            "https://api.exa.ai/search",
            json=payload,
            headers=headers,
            timeout=30,
        )
        response.raise_for_status()
        json_data = response.json()
        return any_to_str(json_data)
    except Exception as e:
        logger.error(f"Exa search failed: {e}")
        return f"Search failed: {str(e)}. Please try again."

# Load environment variables
load_dotenv()

# System prompts for each agent

REQUIREMENTS_ANALYZER_PROMPT = """
You are the Requirements Analyzer Agent for Job Search.

ROLE:
Extract and clarify job search requirements from user input to create optimized search queries.

RESPONSIBILITIES:
- Engage with the user to understand:
  * Desired job titles and roles
  * Required skills and qualifications
  * Preferred locations (remote, hybrid, on-site)
  * Salary expectations
  * Company size and culture preferences
  * Industry preferences
  * Experience level
  * Work authorization status
  * Career goals and priorities

- Analyze user responses to identify:
  * Key search terms and keywords
  * Must-have vs nice-to-have requirements
  * Deal-breakers or constraints
  * Priority factors in job selection

- Generate optimized search queries:
  * Create 3-5 targeted search queries based on user requirements
  * Combine job titles, skills, locations, and key criteria
  * Format queries for maximum relevance

OUTPUT FORMAT:
Provide a comprehensive requirements analysis:
1. User Profile Summary:
   - Job titles of interest
   - Key skills and qualifications
   - Location preferences
   - Salary range
   - Priority factors

2. Search Strategy:
   - List of 3-5 optimized search queries
   - Rationale for each query
   - Expected result types

3. Clarifications Needed (if any):
   - Questions to refine search
   - Missing information

IMPORTANT:
- Always include ALL user responses verbatim in your analysis
- Format search queries clearly for the next agent
- Be specific and actionable in your recommendations
- Ask follow-up questions if requirements are unclear
"""

SEARCH_EXECUTOR_PROMPT = """
You are the Search Executor Agent for Job Search.

ROLE:
Execute job searches using exa_search and analyze results for relevance.

TOOLS:
You have access to the exa_search tool. Use it to find current job listings and career opportunities.

RESPONSIBILITIES:
- Execute searches using queries from the Requirements Analyzer
- Use exa_search for EACH query provided
- Analyze search results for:
  * Job title match
  * Skills alignment
  * Location compatibility
  * Salary range fit
  * Company reputation
  * Role responsibilities
  * Growth opportunities

- Categorize results:
  * Strong Match (80-100% alignment)
  * Good Match (60-79% alignment)
  * Moderate Match (40-59% alignment)
  * Weak Match (<40% alignment)

- For each job listing, extract:
  * Job title and company
  * Location and work arrangement
  * Key requirements
  * Salary range (if available)
  * Application link or contact
  * Match score and reasoning

OUTPUT FORMAT:
Provide structured search results:
1. Search Execution Summary:
   - Queries executed
   - Total results found
   - Distribution by match category

2. Detailed Job Listings (organized by match strength):
   For each job:
   - Company and Job Title
   - Location and Work Type
   - Key Requirements
   - Why it's a match (or not)
   - Match Score (percentage)
   - Application link
   - Source (cite exa_search)

3. Search Insights:
   - Common themes in results
   - Gap analysis (requirements not met)
   - Market observations

INSTRUCTIONS:
- Always use exa_search for EVERY query provided
- Cite exa_search results clearly
- Be objective in match assessment
- Provide actionable insights
"""

RESULTS_CURATOR_PROMPT = """
You are the Results Curator Agent for Job Search.

ROLE:
Filter, organize, and present job search results to the user for decision-making.

RESPONSIBILITIES:
- Review all search results from the Search Executor
- Filter and prioritize based on:
  * Match scores
  * User requirements
  * Application deadlines
  * Job quality indicators

- Organize results into:
  * Top Recommendations (top 3-5 best matches)
  * Strong Alternatives (next 5-10 options)
  * Worth Considering (other relevant matches)

- For top recommendations, provide:
  * Detailed comparison
  * Pros and cons for each
  * Application strategy suggestions
  * Next steps

- Engage user for feedback:
  * Present curated results clearly
  * Ask which jobs interest them
  * Identify what's missing
  * Determine if new search is needed

OUTPUT FORMAT:
Provide a curated job search report:

1. Executive Summary:
   - Total jobs reviewed
   - Number of strong matches
   - Key findings

2. Top Recommendations (detailed):
   For each (max 5):
   - Company & Title
   - Why it's a top match
   - Key highlights
   - Potential concerns
   - Recommendation strength (1-10)
   - Application priority (High/Medium/Low)

3. Strong Alternatives (brief list):
   - Company & Title
   - One-line match summary
   - Match score

4. User Decision Point:
   Ask the user:
   - "Which of these jobs interest you most?"
   - "What's missing from these results?"
   - "Should we refine the search or proceed with applications?"
   - "Any requirements you'd like to adjust?"

5. Next Steps:
   Based on user response, either:
   - Proceed with selected jobs
   - Run new search with adjusted criteria
   - Deep dive into specific opportunities

IMPORTANT:
- Make it easy for users to make decisions
- Be honest about job fit
- Provide clear paths forward
- Always ask for user feedback before concluding
"""

class JobSearchSwarm:
    def __init__(
        self,
        name: str = "AI Job Search Swarm",
        description: str = "An intelligent job search system that finds your ideal role",
        max_loops: int = 1,
        user_name: str = "Job Seeker",
        output_type: str = "json",
    ):
        self.max_loops = max_loops
        self.name = name
        self.description = description
        self.user_name = user_name
        self.output_type = output_type
        
        self.agents = self._initialize_agents()
        self.conversation = Conversation()
        self.exa_search_results = []
        self.search_queries = []
        self.current_iteration = 0
        self.max_iterations = 10  # Prevent infinite loops
        self.search_concluded = False
        
        self.handle_initial_processing()

    def handle_initial_processing(self):
        self.conversation.add(
            role="System",
            content=f"Company: {self.name}\n"
                    f"Description: {self.description}\n"
                    f"Mission: Find the perfect job match for {self.user_name}"
        )

    def _initialize_agents(self) -> List[Agent]:
        return [
            Agent(
                agent_name="Sarah-Requirements-Analyzer",
                agent_description="Analyzes user requirements and creates optimized job search queries.",
                system_prompt=REQUIREMENTS_ANALYZER_PROMPT,
                max_loops=self.max_loops,
                dynamic_temperature_enabled=True,
                output_type="final",
            ),
            Agent(
                agent_name="David-Search-Executor",
                agent_description="Executes job searches and analyzes results for relevance.",
                system_prompt=SEARCH_EXECUTOR_PROMPT,
                max_loops=self.max_loops,
                dynamic_temperature_enabled=True,
                output_type="final",
            ),
            Agent(
                agent_name="Lisa-Results-Curator",
                agent_description="Curates and presents job results for user decision-making.",
                system_prompt=RESULTS_CURATOR_PROMPT,
                max_loops=self.max_loops,
                dynamic_temperature_enabled=True,
                output_type="final",
            ),
        ]

    def find_agent_by_name(self, name: str) -> Agent:
        for agent in self.agents:
            if name in agent.agent_name:
                return agent
        return None

    def analyze_requirements(self, user_input: str):
        """Phase 1: Analyze user requirements and generate search queries"""
        sarah_agent = self.find_agent_by_name("Requirements-Analyzer")
        
        sarah_output = sarah_agent.run(
            f"User Input: {user_input}\n\n"
            f"Conversation History: {self.conversation.get_str()}\n\n"
            f"Analyze the user's job search requirements and generate 3-5 optimized search queries. "
            f"If information is unclear, ask clarifying questions."
        )
        
        self.conversation.add(
            role="Requirements-Analyzer", content=sarah_output
        )
        
        # Extract search queries from Sarah's output
        self.search_queries = self._extract_search_queries(sarah_output)
        
        return sarah_output

    def _extract_search_queries(self, analyzer_output: str) -> List[str]:
        """Extract search queries from Requirements Analyzer output"""
        queries = []
        lines = analyzer_output.split('\n')
        
        # Look for lines that appear to be search queries
        for line in lines:
            line = line.strip()
            # Simple heuristic: lines with certain keywords or patterns
            if any(keyword in line.lower() for keyword in ['query:', 'search:', 'query']):
                # Extract the actual query
                if ':' in line:
                    query = line.split(':', 1)[1].strip()
                    if query and len(query) > 10:
                        queries.append(query)
        
        # If no queries found, create default ones based on common patterns
        if not queries:
            logger.warning("No explicit queries found, generating fallback queries")
            queries = [
                "software engineer jobs remote",
                "data scientist positions",
                "product manager opportunities"
            ]
        
        return queries[:5]  # Limit to 5 queries

    def execute_searches(self):
        """Phase 2: Execute searches using exa_search and analyze results"""
        # Execute exa_search for each query
        self.exa_search_results = []
        for query in self.search_queries:
            result = exa_search(query)
            self.exa_search_results.append({
                "query": query,
                "exa_result": result
            })
        
        # Pass results to Search Executor agent
        david_agent = self.find_agent_by_name("Search-Executor")
        
        # Build exa context
        exa_context = "\n\n[Exa Search Results]\n"
        for item in self.exa_search_results:
            exa_context += f"Query: {item['query']}\nResults: {item['exa_result']}\n\n"
        
        david_output = david_agent.run(
            f"Conversation History: {self.conversation.get_str()}\n\n"
            f"{exa_context}\n"
            f"Analyze these job search results. Categorize each job by match strength and provide detailed analysis."
        )
        
        self.conversation.add(
            role="Search-Executor", content=david_output
        )
        
        return david_output

    def curate_results(self) -> str:
        """Phase 3: Curate results and get user feedback"""
        lisa_agent = self.find_agent_by_name("Results-Curator")
        
        lisa_output = lisa_agent.run(
            f"Conversation History: {self.conversation.get_str()}\n\n"
            f"Curate the job search results, present top recommendations, and ask the user for feedback. "
            f"Determine if we should continue searching or if the user has found suitable options."
        )
        
        self.conversation.add(
            role="Results-Curator", content=lisa_output
        )
        
        return lisa_output

    def end(self) -> tuple[bool, str]:
        """
        Conclude the job search without user interaction.
        
        Returns:
            tuple[bool, str]: (needs_refinement, user_feedback)
        """
        return False, "Search completed successfully."

    def run(self, initial_user_input: str):
        """
        Run the job search swarm with continuous optimization.
        
        Args:
            initial_user_input: User's initial job search requirements
        """
        self.conversation.add(role=self.user_name, content=initial_user_input)
        
        user_input = initial_user_input
        
        while not self.search_concluded and self.current_iteration < self.max_iterations:
            self.current_iteration += 1
            logger.info(f"Starting search iteration {self.current_iteration}")
            
            # Phase 1: Analyze requirements
            print(f"\n{'='*60}")
            print(f"ITERATION {self.current_iteration} - ANALYZING REQUIREMENTS")
            print(f"{'='*60}\n")
            self.analyze_requirements(user_input)

            
            # Phase 2: Execute searches
            print(f"\n{'='*60}")
            print(f"ITERATION {self.current_iteration} - EXECUTING JOB SEARCHES")
            print(f"{'='*60}\n")
            self.execute_searches()
            
            # Phase 3: Curate and present results
            print(f"\n{'='*60}")
            print(f"ITERATION {self.current_iteration} - CURATING RESULTS")
            print(f"{'='*60}\n")
            self.curate_results()
            
            # Phase 4: Get user feedback
            needs_refinement, user_feedback = self.end()
            
            # Add user feedback to conversation
            self.conversation.add(
                role=self.user_name, 
                content=f"User Feedback: {user_feedback}"
            )
            
            # Check if we should continue
            if not needs_refinement:
                self.search_concluded = True
                print(f"\n{'='*60}")
                print("SEARCH CONCLUDED - USER SATISFIED WITH RESULTS")
                print(f"{'='*60}\n")
            else:
                # In production, get new user input here
                print(f"\n{'='*60}")
                print("SEARCH REQUIRES REFINEMENT")
                print(f"{'='*60}\n")
                # For demo, we'll stop after first iteration
                self.search_concluded = True
        
        # Return formatted conversation history
        return history_output_formatter(
            self.conversation, type=self.output_type
        )

def main():
    """Main entry point for job search swarm"""
    
    # Example 1: Pre-filled user requirements (for testing)
    user_requirements = """
    I'm looking for a senior software engineer position with the following requirements:
    - Job Title: Senior Software Engineer or Staff Engineer
    - Skills: Python, distributed systems, cloud architecture (AWS/GCP), Kubernetes
    - Location: Remote (US-based) or San Francisco Bay Area
    - Salary: $180k - $250k
    - Company: Mid-size to large tech companies, prefer companies with strong engineering culture
    - Experience Level: 7+ years
    - Industry: SaaS, Cloud Infrastructure, or Developer Tools
    - Work Authorization: US Citizen
    - Priorities: Technical challenges, work-life balance, remote flexibility, equity upside
    - Deal-breakers: No pure management roles, no strict return-to-office policies
    """
    
    # Initialize the swarm
    job_search_swarm = JobSearchSwarm(
        name="AI-Powered Job Search Engine",
        description="Intelligent job search system that continuously refines results until the perfect match is found",
        user_name="Job Seeker",
        output_type="json",
        max_loops=1,
    )
    
    # Run the swarm
    print("\n" + "="*60)
    print("INITIALIZING JOB SEARCH SWARM")
    print("="*60 + "\n")
    
    job_search_swarm.run(initial_user_input=user_requirements)

if __name__ == "__main__":
    main()

Upon execution, the swarm will:

1. Analyze the provided user_requirements.
2. Generate and execute search queries using Exa.
3. Curate and present the results in a structured format, including top recommendations and a prompt for user feedback.

The output will be printed to the console, showing the progression of the agents through each phase of the job search.

Workflow Stages

The JobSearchSwarm processes the job search through a continuous, iterative workflow:

1. Phase 1: Analyze Requirements (analyze_requirements): The Sarah-Requirements-Analyzer agent processes the user's input and conversation history to extract job criteria and generate optimized search queries.
2. Phase 2: Execute Searches (execute_searches): The David-Search-Executor agent takes the generated queries, uses the exa_search tool to find job listings, and analyzes their relevance against the user's requirements.
3. Phase 3: Curate Results (curate_results): The Lisa-Results-Curator agent reviews, filters, and organizes the search results, presenting top recommendations and asking for user feedback to guide further iterations.
4. Phase 4: Get User Feedback (end): In a full implementation, this stage gathers explicit user feedback to determine if the search needs refinement or can be concluded. For demonstration, this is a simulated step.

The swarm continues these phases in a loop until the search_concluded flag is set to True or max_iterations is reached.

Customization

You can customize the JobSearchSwarm by modifying the JobSearchSwarm class parameters or the agents' prompts:

• name and description: Customize the swarm's identity.
• user_name: Define the name of the user interacting with the swarm.
• output_type: Specify the desired output format for the conversation history (e.g., "json" or "list").
• max_loops: Control the number of internal reasoning iterations each agent performs (set during agent initialization).
• system_prompt: Modify the REQUIREMENTS_ANALYZER_PROMPT, SEARCH_EXECUTOR_PROMPT, and RESULTS_CURATOR_PROMPT to refine agent behavior and output.
• max_iterations: Limit the total number of search cycles the swarm performs.

Best Practices

To get the most out of the AI Job Search Swarm:

• Provide Clear Requirements: Start with a detailed and unambiguous initial_user_input to help the Requirements Analyzer generate effective queries.
• Iterate and Refine: In a live application, leverage the user feedback loop to continuously refine search criteria and improve result relevance.
• Monitor Agent Outputs: Regularly review the outputs from each agent to ensure they are performing as expected and to identify areas for prompt improvement.
• Manage API Usage: Be mindful of your Exa API key usage, especially when experimenting with max_iterations or a large number of search queries.

Limitations

• Prompt Engineering Dependency: The quality of the search results heavily depends on the clarity and effectiveness of the agent system_prompts and the initial user input.
• Exa Search Scope: The exa_search tool's effectiveness is tied to the breadth and depth of Exa's indexed web content.
• Iteration Control: The current end method in examples/demos/apps/job_finding.py is simplified for demonstration. A robust production system would require a more sophisticated user interaction mechanism to determine when to stop or refine the search.
• Verification Needed: All AI-generated outputs, including job matches and summaries, should be independently verified by the user.