pip install -U praxos-python

# Replace with your Praxos API key
API_KEY = "prx_your_api_key_here"

from praxos_python import SyncClient

# Initialize the client with your API key
client = SyncClient(api_key=API_KEY)

# Validate your API key
try:
    is_valid = client.validate_api_key()
    print(f"API key is valid: {is_valid}")
except Exception as e:
    print(f"API key validation failed: {e}")

# Create a new environment or get an existing one
env = client.create_environment(name="my-quickstart-environment")

# Alternative: Get an existing environment
# env = client.get_environment(name="my-quickstart-environment")

from praxos_python.types import Message

# Create conversation messages
messages = [
    Message(role="user", content="What are our Q4 sales targets for enterprise clients?"),
    Message(role="assistant", content="Q4 targets are $2.5M with focus on Fortune 500 companies."),
    Message(role="user", content="What about our current pipeline?"),
    Message(role="assistant", content="Current pipeline shows $1.8M in qualified opportunities.")
]

conversation_source = env.add_conversation(
    messages=messages,
    name="Sales Planning Discussion",
    description="Q4 sales targets and pipeline review"
)

# Sample business data
business_data = {
    "customers": [
        {"name": "TechCorp Inc", "tier": "enterprise", "revenue": 250000, "industry": "Technology"},
        {"name": "Global Manufacturing", "tier": "premium", "revenue": 180000, "industry": "Manufacturing"},
        {"name": "StartupXYZ", "tier": "basic", "revenue": 25000, "industry": "Software"}
    ],
    "quarterly_targets": {
        "Q4_2023": {"target": 2500000, "current": 1800000, "forecast": 2200000}
    }
}

data_source = env.add_business_data(
    data=business_data,
    name="Customer and Sales Data",
    description="Q4 customer information and sales targets"
)

import time

# Check status of both sources
sources = [conversation_source, data_source]

for source in sources:
    print(f"Checking status for: {source.name}")
    
    # Wait for indexing to complete
    while source.get_status() != "success":
        status = source.get_status()
        print(f"Status: {status}...")
        if status == "failed":
            print("Indexing failed!")
            break
        time.sleep(2)
    
    print(f"✅ {source.name} is ready!")

# Ask natural language questions about your data
results = env.intelligent_search("What are our highest revenue customers?")

print("=== Search Results ===")
for hit in results["hits"]:
    print(f"Score: {hit['score']:.2f}")
    print(f"Content: {hit['sentence']}")
    print("---")

# Access AI analysis insights
analysis = results["intelligent_analysis"]
print(f"\n=== AI Analysis ===")
print(f"Query Intent: {analysis.get('intent', 'unknown')}")
print(f"Strategies Used: {analysis['strategies_used']}")

# Check for automatic analytical operations
if "toolset_analysis" in results:
    print(f"\n=== Analytical Results ===")
    for operation in results["toolset_analysis"]:
        if operation['operation'] == 'find_maximum':
            print(f"Highest Revenue: ${operation['max_value']:,.2f}")
            print(f"Customer: {operation['max_item']}")

# Ask about sales targets
results = env.intelligent_search("What are our Q4 sales targets and current progress?")

# Ask analytical questions
results = env.intelligent_search("What's the average customer revenue by industry?")

# Ask temporal questions
results = env.intelligent_search("Show me enterprise customers with revenue above $200k")

# The AI automatically:
# - Understands the query intent
# - Identifies relevant data types
# - Selects optimal search strategies  
# - Performs calculations when needed

# Extract customer entities
customers = env.extract_items(
    schema="Customer",  # Assumes you have a Customer schema in your ontology
    top_k=50
)

# Extract revenue amounts
revenue_data = env.extract_literals(
    literal_type="CurrencyAmountType",
    mode="full_entities",
    top_k=100
)

print(f"Found {len(customers)} customers and {len(revenue_data['items'])} revenue records")

# Simple context query  
context = env.get_context("what are the sales targets?")
print(f"Context: {context.sentence}")
print(f"Relevance Score: {context.score}")