Code Examples

pip install medisearch-client

from medisearch_client import MediSearchClient
import uuid

# Initialize client with your API key
api_key = "your_api_key"
client = MediSearchClient(api_key=api_key)

# Generate a unique conversation ID
conversation_id = str(uuid.uuid4())

# Send a query
responses = client.send_message(
    conversation=["What are the symptoms of COVID-19?"],
    conversation_id=conversation_id
)

# Process responses
for response in responses:
    if response["event"] == "llm_response":
        # The AI-generated answer
        print("Answer:", response["data"])
    elif response["event"] == "articles":
        # Supporting articles and citations
        articles = response["data"]
        print(f"Found {len(articles)} supporting articles")
        for i, article in enumerate(articles[:3], 1):
            print(f"{i}. {article['title']} ({article.get('year', 'N/A')})")

# Enable streaming with should_stream_response=True
for response in client.send_message(
    conversation=["What are the treatment options for COVID-19?"],
    conversation_id=str(uuid.uuid4()),
    should_stream_response=True
):
    if response["event"] == "llm_response":
        # Print each chunk as it arrives
        print(response["data"], end="", flush=True)
    elif response["event"] == "articles":
        print(f"Based on {len(response['data'])} medical sources")

# Initial query
conversation = ["What are the symptoms of type 2 diabetes?"]
conversation_id = str(uuid.uuid4())  # Use the same ID for the entire conversation

# Send first message
first_responses = client.send_message(
    conversation=conversation,
    conversation_id=conversation_id
)

# Get the AI's response to include in conversation history
ai_response = None
for response in first_responses:
    if response["event"] == "llm_response":
        ai_response = response["data"]

# Add the AI's response and a follow-up question to the conversation
if ai_response:
    conversation.append(ai_response)
    conversation.append("How is it diagnosed?")
    
    # Send follow-up with complete conversation history
    followup_responses = client.send_message(
        conversation=conversation,
        conversation_id=conversation_id  # Same conversation ID
    )

from medisearch_client import Filters, Settings

# Create filters for customized search
filters = Filters(
    # Specify which sources to search
    sources=[
        "scientificArticles",
        "internationalHealthGuidelines"
    ],
    # Limit to recent publications
    year_start=2020,
    # Only include high-quality sources
    only_high_quality=True,
    # Specific article types
    article_types=[
        "metaAnalysis",
        "clinicalTrials"
    ]
)

# Apply filters through settings
settings = Settings(
    language="English",
    filters=filters,
    model_type="pro"
)

# Send message with custom settings
responses = client.send_message(
    conversation=["What are the latest treatments for Alzheimer's?"],
    conversation_id=str(uuid.uuid4()),
    settings=settings
)

from medisearch_client import ResponseHandler

# Define handler functions
def handle_llm_response(response):
    print(response["data"], end="", flush=True)

def handle_articles(response):
    articles = response["data"]
    print(f"Found {len(articles)} relevant medical articles")
    for i, article in enumerate(articles[:3], 1):
        print(f"{i}. {article['title']} ({article.get('year', 'N/A')})")

def handle_error(response):
    error_code = response["data"]
    print(f"Error occurred: {error_code}")

# Create response handler
handler = ResponseHandler(
    on_llm_response=handle_llm_response,
    on_articles=handle_articles,
    on_error=handle_error
)

# Send message with the response handler
client.send_message(
    conversation=["What are recent advances in breast cancer treatment?"],
    conversation_id=str(uuid.uuid4()),
    response_handler=handler
)

import asyncio
from medisearch_client import MediSearchClient

async def main():
    client = MediSearchClient(api_key="your_api_key")
    
    # Async request
    responses = await client.send_message_async(
        conversation=["What are the cardiovascular effects of COVID-19?"],
        conversation_id=str(uuid.uuid4())
    )
    
    # Process responses
    for response in responses:
        if response["event"] == "llm_response":
            print(f"Answer: {response['data']}")

# Run the async function
asyncio.run(main())

from medisearch_client import Settings

# Create settings with Spanish language preference
settings = Settings(language="Spanish")

# Send query in Spanish
client.send_message(
    conversation=["¿Cuáles son los síntomas de la diabetes?"],
    conversation_id=str(uuid.uuid4()),
    settings=settings,
    response_handler=handler  # Using the previously defined handler
)

from medisearch_client import Settings

# Create settings that specify the Pro model
pro_settings = Settings(
    language="English",
    model_type="pro"  # Explicitly request the Pro model
)

# Complex medical query that benefits from Pro model
responses = client.send_message(
    conversation=[
        "What is the current understanding of the role of microbiome in inflammatory bowel disease pathogenesis?"
    ],
    conversation_id=str(uuid.uuid4()),
    settings=pro_settings
)

from medisearch_client import Settings

# Create settings with a custom system prompt focused on style and simplicity
custom_settings = Settings(
    language="English",
    model_type="pro",
    system_prompt="""Please provide all responses in simple, consumer-friendly language.
Avoid medical jargon when possible, and when medical terms must be used, define them clearly.
Keep answers concise, use bullet points for clarity, and prioritize practical advice.
Limit responses to 3-4 short paragraphs maximum."""
)

# Send a query with the consumer-friendly system prompt
responses = client.send_message(
    conversation=["What's the current approach to managing hypertension in patients with diabetes?"],
    conversation_id=str(uuid.uuid4()),
    settings=custom_settings
)