Skip to main content

​
Competitive Intelligence Dashboard

This example demonstrates how to analyze multiple competitors simultaneously using ConcurrentWorkflow - perfect for parallel, independent tasks.

​
Step 1: Setup

import requests
import os

API_BASE_URL = "https://api.swarms.world"
API_KEY = os.environ.get("SWARMS_API_KEY", "your_api_key_here")

headers = {
    "x-api-key": API_KEY,
    "Content-Type": "application/json"
}

​
Step 2: Define Parallel Competitor Analysis

Create one analyst agent per competitor - all run simultaneously: 
def analyze_competitors(competitors: list[str], industry: str) -> dict:
    """Analyze multiple competitors in parallel."""

    # Create one agent per competitor
    agents = []
    for competitor in competitors:
        agents.append({
            "agent_name": f"{competitor} Analyst",
            "description": f"Analyzes {competitor}'s market position",
            "system_prompt": f"""You are a competitive intelligence analyst for {competitor}.

Analyze across these dimensions:
1. COMPANY OVERVIEW - Founded, size, funding, leadership
2. PRODUCTS - Core offerings, recent launches, unique features
3. MARKET POSITION - Target segments, market share, partnerships
4. PRICING - Model, price points, enterprise approach
5. STRENGTHS & WEAKNESSES - Top 3 of each

Be specific with facts and numbers.""",
            "model_name": "gpt-4o",
            "max_loops": 1,
            "temperature": 0.3
        })

    swarm_config = {
        "name": f"Competitive Intelligence: {industry}",
        "description": "Parallel competitor analysis",
        "swarm_type": "ConcurrentWorkflow",
        "task": f"Analyze market position in the {industry} industry for: {', '.join(competitors)}",
        "agents": agents,
        "max_loops": 1
    }

    response = requests.post(
        f"{API_BASE_URL}/v1/swarm/completions",
        headers=headers,
        json=swarm_config,
        timeout=180
    )

    return response.json()

​
Step 3: Run Parallel Analysis

# Define competitors to analyze
competitors = ["OpenAI", "Anthropic", "Google DeepMind", "Cohere", "Mistral AI"]

# Run parallel analysis
result = analyze_competitors(competitors, "AI/LLM Providers")

# Display results
execution_time = result.get("execution_time", 0)
num_competitors = len(result.get("output", []))

print(f"Analyzed {num_competitors} competitors in {execution_time:.1f}s")
print(f"Sequential would take ~{execution_time * num_competitors:.1f}s ({num_competitors}x slower)")

for output in result.get("output", []):
    competitor = output["role"].replace(" Analyst", "")
    content = output["content"]

    print(f"\n{'='*50}")
    print(f"{competitor.upper()}")
    print(f"{'='*50}")
    print(content[:600] + "...")

print(f"\nTotal cost: ${result['usage']['billing_info']['total_cost']:.4f}")
Expected Output: 
Analyzed 5 competitors in 28.4s
Sequential would take ~142.0s (5x slower)

==================================================
OPENAI
==================================================
COMPANY OVERVIEW
- Founded: 2015, San Francisco
- Employees: ~1,500+
- Valuation: $80B+ (2024)
- Leadership: Sam Altman (CEO)

PRODUCTS
- Core: GPT-4, ChatGPT, DALL-E, Whisper
- Recent: GPT-4 Turbo, Custom GPTs, GPT Store...

==================================================
ANTHROPIC
==================================================
COMPANY OVERVIEW
- Founded: 2021, San Francisco
- Employees: ~500...

Total cost: $0.2156
ConcurrentWorkflow runs all agents in parallel. Use it when tasks are independent and don’t need each other’s output. You pay the same tokens whether sequential or concurrent, but finish faster.