aishastr.live

Agentic Design Pattern: Api Gateway

Written by

admin

in


    """
Pattern 156: API Gateway Pattern

Description:
    Centralized gateway for managing API requests to multiple agents/services.
    Provides routing, authentication, rate limiting, monitoring, and unified
    interface for diverse backend services.

Components:
    - Request router
    - Authentication & authorization
    - Rate limiting
    - Load balancing
    - Monitoring & logging
    - Response transformation
    - Error handling

Use Cases:
    - Multi-agent systems
    - Microservices architecture
    - API versioning
    - Security enforcement
    - Traffic management

Benefits:
    - Centralized control
    - Security enforcement
    - Traffic management
    - Simplified client interface
    - Monitoring visibility

Trade-offs:
    - Single point of failure
    - Added latency
    - Complexity overhead
    - Gateway bottleneck potential

LangChain Implementation:
    Uses routing logic, rate limiting, authentication checks,
    and multiple LLM backends with unified interface
"""

import os
import time
from typing import Dict, Any, Optional, Callable, List
from dataclasses import dataclass, field
from datetime import datetime, timedelta
from enum import Enum
from collections import defaultdict
from dotenv import load_dotenv

from langchain_openai import ChatOpenAI
from langchain_core.prompts import ChatPromptTemplate
from langchain_core.output_parsers import StrOutputParser

load_dotenv()


class ServiceType(Enum):
    """Types of backend services"""
    CHAT = "chat"
    COMPLETION = "completion"
    ANALYSIS = "analysis"
    TRANSLATION = "translation"
    SUMMARIZATION = "summarization"


class AuthLevel(Enum):
    """Authentication levels"""
    PUBLIC = "public"
    AUTHENTICATED = "authenticated"
    PREMIUM = "premium"
    ADMIN = "admin"


@dataclass
class Route:
    """API route configuration"""
    path: str
    service_type: ServiceType
    handler: Callable
    auth_level: AuthLevel = AuthLevel.PUBLIC
    rate_limit: int = 60  # requests per minute
    timeout: float = 30.0  # seconds


@dataclass
class ServiceConfig:
    """Backend service configuration"""
    name: str
    service_type: ServiceType
    model_name: str
    enabled: bool = True
    weight: int = 1  # For load balancing
    max_concurrent: int = 10


@dataclass
class Request:
    """Gateway request"""
    path: str
    method: str
    payload: Dict[str, Any]
    headers: Dict[str, str] = field(default_factory=dict)
    user_id: Optional[str] = None
    api_key: Optional[str] = None
    timestamp: datetime = field(default_factory=datetime.now)


@dataclass
class Response:
    """Gateway response"""
    status_code: int
    data: Any
    headers: Dict[str, str] = field(default_factory=dict)
    execution_time: float = 0.0
    service_name: Optional[str] = None
    error: Optional[str] = None


@dataclass
class RateLimitInfo:
    """Rate limit tracking"""
    requests: List[datetime] = field(default_factory=list)
    limit: int = 60
    window: int = 60  # seconds


class APIGateway:
    """Centralized API gateway for agent services"""
    
    def __init__(self):
        """Initialize API gateway"""
        self.routes: Dict[str, Route] = {}
        self.services: Dict[str, ServiceConfig] = {}
        self.rate_limits: Dict[str, RateLimitInfo] = defaultdict(
            lambda: RateLimitInfo()
        )
        self.api_keys: Dict[str, AuthLevel] = {}
        self.request_log: List[Dict[str, Any]] = []
        
        # Initialize services
        self._initialize_services()
        
        # Set up routes
        self._setup_routes()
    
    def _initialize_services(self):
        """Initialize backend services"""
        self.services = {
            "chat_service": ServiceConfig(
                name="chat_service",
                service_type=ServiceType.CHAT,
                model_name="gpt-3.5-turbo"
            ),
            "analysis_service": ServiceConfig(
                name="analysis_service",
                service_type=ServiceType.ANALYSIS,
                model_name="gpt-3.5-turbo"
            ),
            "translation_service": ServiceConfig(
                name="translation_service",
                service_type=ServiceType.TRANSLATION,
                model_name="gpt-3.5-turbo"
            ),
        }
        
        # Initialize LLMs for services
        self.llms = {
            name: ChatOpenAI(model=config.model_name, temperature=0.7)
            for name, config in self.services.items()
        }
    
    def _setup_routes(self):
        """Set up API routes"""
        self.routes = {
            "/chat": Route(
                path="/chat",
                service_type=ServiceType.CHAT,
                handler=self._handle_chat,
                auth_level=AuthLevel.AUTHENTICATED,
                rate_limit=30
            ),
            "/analyze": Route(
                path="/analyze",
                service_type=ServiceType.ANALYSIS,
                handler=self._handle_analysis,
                auth_level=AuthLevel.PREMIUM,
                rate_limit=10
            ),
            "/translate": Route(
                path="/translate",
                service_type=ServiceType.TRANSLATION,
                handler=self._handle_translation,
                auth_level=AuthLevel.AUTHENTICATED,
                rate_limit=20
            ),
            "/summarize": Route(
                path="/summarize",
                service_type=ServiceType.SUMMARIZATION,
                handler=self._handle_summarization,
                auth_level=AuthLevel.AUTHENTICATED,
                rate_limit=15
            ),
        }
    
    def register_api_key(self, api_key: str, auth_level: AuthLevel):
        """Register an API key with authentication level"""
        self.api_keys[api_key] = auth_level
    
    def authenticate(self, request: Request) -> tuple[bool, Optional[str]]:
        """Authenticate request"""
        # Get route
        route = self.routes.get(request.path)
        if not route:
            return False, "Route not found"
        
        # Check authentication level
        if route.auth_level == AuthLevel.PUBLIC:
            return True, None
        
        # Check API key
        if not request.api_key:
            return False, "API key required"
        
        user_level = self.api_keys.get(request.api_key)
        if not user_level:
            return False, "Invalid API key"
        
        # Check authorization level
        required_levels = {
            AuthLevel.AUTHENTICATED: [AuthLevel.AUTHENTICATED, AuthLevel.PREMIUM, AuthLevel.ADMIN],
            AuthLevel.PREMIUM: [AuthLevel.PREMIUM, AuthLevel.ADMIN],
            AuthLevel.ADMIN: [AuthLevel.ADMIN]
        }
        
        if user_level not in required_levels.get(route.auth_level, []):
            return False, f"Insufficient permissions. Requires {route.auth_level.value}"
        
        return True, None
    
    def check_rate_limit(self, request: Request) -> tuple[bool, Optional[str]]:
        """Check if request exceeds rate limit"""
        route = self.routes.get(request.path)
        if not route:
            return True, None
        
        # Get rate limit info for user
        key = request.api_key or request.user_id or "anonymous"
        rate_info = self.rate_limits[key]
        
        # Update limit if different from current
        if rate_info.limit != route.rate_limit:
            rate_info.limit = route.rate_limit
        
        # Clean old requests outside window
        now = datetime.now()
        cutoff = now - timedelta(seconds=rate_info.window)
        rate_info.requests = [
            req_time for req_time in rate_info.requests
            if req_time > cutoff
        ]
        
        # Check limit
        if len(rate_info.requests) >= rate_info.limit:
            return False, f"Rate limit exceeded. Max {rate_info.limit} requests per minute"
        
        # Add current request
        rate_info.requests.append(now)
        return True, None
    
    def route_request(self, request: Request) -> Response:
        """Route and process request through gateway"""
        start_time = time.time()
        
        # Log request
        self._log_request(request)
        
        # Authenticate
        auth_ok, auth_error = self.authenticate(request)
        if not auth_ok:
            return Response(
                status_code=401,
                data=None,
                error=auth_error,
                execution_time=time.time() - start_time
            )
        
        # Check rate limit
        rate_ok, rate_error = self.check_rate_limit(request)
        if not rate_ok:
            return Response(
                status_code=429,
                data=None,
                error=rate_error,
                execution_time=time.time() - start_time
            )
        
        # Get route
        route = self.routes.get(request.path)
        if not route:
            return Response(
                status_code=404,
                data=None,
                error=f"Route not found: {request.path}",
                execution_time=time.time() - start_time
            )
        
        # Select service
        service = self._select_service(route.service_type)
        if not service:
            return Response(
                status_code=503,
                data=None,
                error="No available service",
                execution_time=time.time() - start_time
            )
        
        # Execute request
        try:
            result = route.handler(request, service)
            execution_time = time.time() - start_time
            
            return Response(
                status_code=200,
                data=result,
                execution_time=execution_time,
                service_name=service.name
            )
            
        except Exception as e:
            execution_time = time.time() - start_time
            return Response(
                status_code=500,
                data=None,
                error=str(e),
                execution_time=execution_time,
                service_name=service.name
            )
    
    def _select_service(self, service_type: ServiceType) -> Optional[ServiceConfig]:
        """Select service based on type and load balancing"""
        # Find services matching type
        matching = [
            s for s in self.services.values()
            if s.service_type == service_type and s.enabled
        ]
        
        if not matching:
            return None
        
        # Simple weighted random selection
        import random
        weights = [s.weight for s in matching]
        return random.choices(matching, weights=weights)[0]
    
    def _handle_chat(self, request: Request, service: ServiceConfig) -> str:
        """Handle chat request"""
        llm = self.llms[service.name]
        
        prompt = ChatPromptTemplate.from_messages([
            ("system", "You are a helpful AI assistant."),
            ("user", "{message}")
        ])
        
        chain = prompt | llm | StrOutputParser()
        return chain.invoke({"message": request.payload.get("message", "")})
    
    def _handle_analysis(self, request: Request, service: ServiceConfig) -> str:
        """Handle analysis request"""
        llm = self.llms[service.name]
        
        prompt = ChatPromptTemplate.from_messages([
            ("system", "You are an expert analyst. Provide detailed analysis."),
            ("user", "Analyze the following:\n\n{text}")
        ])
        
        chain = prompt | llm | StrOutputParser()
        return chain.invoke({"text": request.payload.get("text", "")})
    
    def _handle_translation(self, request: Request, service: ServiceConfig) -> str:
        """Handle translation request"""
        llm = self.llms[service.name]
        
        prompt = ChatPromptTemplate.from_messages([
            ("system", "You are a professional translator."),
            ("user", "Translate the following from {source} to {target}:\n\n{text}")
        ])
        
        chain = prompt | llm | StrOutputParser()
        return chain.invoke({
            "source": request.payload.get("source_lang", "auto"),
            "target": request.payload.get("target_lang", "English"),
            "text": request.payload.get("text", "")
        })
    
    def _handle_summarization(self, request: Request, service: ServiceConfig) -> str:
        """Handle summarization request"""
        llm = self.llms[service.name]
        
        prompt = ChatPromptTemplate.from_messages([
            ("system", "You are an expert at creating concise summaries."),
            ("user", "Summarize the following in {max_words} words:\n\n{text}")
        ])
        
        chain = prompt | llm | StrOutputParser()
        return chain.invoke({
            "max_words": request.payload.get("max_words", 100),
            "text": request.payload.get("text", "")
        })
    
    def _log_request(self, request: Request):
        """Log request for monitoring"""
        self.request_log.append({
            "timestamp": request.timestamp.isoformat(),
            "path": request.path,
            "method": request.method,
            "user_id": request.user_id,
            "api_key_hash": hash(request.api_key) if request.api_key else None
        })
    
    def get_metrics(self) -> Dict[str, Any]:
        """Get gateway metrics"""
        now = datetime.now()
        recent = now - timedelta(minutes=5)
        
        recent_requests = [
            log for log in self.request_log
            if datetime.fromisoformat(log["timestamp"]) > recent
        ]
        
        # Count by path
        by_path = defaultdict(int)
        for log in recent_requests:
            by_path[log["path"]] += 1
        
        # Rate limit status
        rate_status = {}
        for key, info in self.rate_limits.items():
            rate_status[key] = {
                "current": len(info.requests),
                "limit": info.limit,
                "usage_pct": len(info.requests) / info.limit * 100
            }
        
        return {
            "total_requests_5min": len(recent_requests),
            "requests_by_path": dict(by_path),
            "rate_limit_status": rate_status,
            "active_services": len([s for s in self.services.values() if s.enabled]),
            "total_services": len(self.services)
        }


def demonstrate_api_gateway():
    """Demonstrate API gateway pattern"""
    print("=" * 80)
    print("API GATEWAY PATTERN DEMONSTRATION")
    print("=" * 80)
    
    # Initialize gateway
    gateway = APIGateway()
    
    # Register API keys
    gateway.register_api_key("free_user_key", AuthLevel.AUTHENTICATED)
    gateway.register_api_key("premium_user_key", AuthLevel.PREMIUM)
    gateway.register_api_key("admin_key", AuthLevel.ADMIN)
    
    # Example 1: Successful authenticated request
    print("\n" + "="*80)
    print("EXAMPLE 1: Successful Chat Request (Authenticated)")
    print("="*80)
    request = Request(
        path="/chat",
        method="POST",
        payload={"message": "What is machine learning?"},
        api_key="free_user_key",
        user_id="user123"
    )
    response = gateway.route_request(request)
    print(f"Status: {response.status_code}")
    print(f"Service: {response.service_name}")
    print(f"Time: {response.execution_time:.2f}s")
    print(f"Response: {response.data[:200] if response.data else response.error}...")
    
    # Example 2: Premium service request
    print("\n" + "="*80)
    print("EXAMPLE 2: Analysis Request (Premium)")
    print("="*80)
    request = Request(
        path="/analyze",
        method="POST",
        payload={"text": "The market showed strong growth in Q1 with a 15% increase in sales."},
        api_key="premium_user_key"
    )
    response = gateway.route_request(request)
    print(f"Status: {response.status_code}")
    print(f"Service: {response.service_name}")
    print(f"Response: {response.data[:200] if response.data else response.error}...")
    
    # Example 3: Authentication failure
    print("\n" + "="*80)
    print("EXAMPLE 3: Authentication Failure")
    print("="*80)
    request = Request(
        path="/analyze",
        method="POST",
        payload={"text": "Test"},
        api_key="free_user_key"  # Not premium
    )
    response = gateway.route_request(request)
    print(f"Status: {response.status_code}")
    print(f"Error: {response.error}")
    
    # Example 4: Rate limiting
    print("\n" + "="*80)
    print("EXAMPLE 4: Rate Limiting")
    print("="*80)
    # Make multiple requests quickly
    for i in range(3):
        request = Request(
            path="/chat",
            method="POST",
            payload={"message": f"Request {i+1}"},
            api_key="free_user_key"
        )
        response = gateway.route_request(request)
        print(f"Request {i+1}: Status {response.status_code} - {response.error if response.error else 'Success'}")
    
    # Example 5: Translation service
    print("\n" + "="*80)
    print("EXAMPLE 5: Translation Service")
    print("="*80)
    request = Request(
        path="/translate",
        method="POST",
        payload={
            "text": "Hello, how are you?",
            "source_lang": "English",
            "target_lang": "Spanish"
        },
        api_key="free_user_key"
    )
    response = gateway.route_request(request)
    print(f"Status: {response.status_code}")
    print(f"Translation: {response.data if response.data else response.error}")
    
    # Example 6: Gateway metrics
    print("\n" + "="*80)
    print("EXAMPLE 6: Gateway Metrics")
    print("="*80)
    metrics = gateway.get_metrics()
    print("Gateway Metrics:")
    print(f"  Total requests (5 min): {metrics['total_requests_5min']}")
    print(f"  Requests by path: {metrics['requests_by_path']}")
    print(f"  Active services: {metrics['active_services']}/{metrics['total_services']}")
    
    # Example 7: Route not found
    print("\n" + "="*80)
    print("EXAMPLE 7: Route Not Found")
    print("="*80)
    request = Request(
        path="/nonexistent",
        method="POST",
        payload={},
        api_key="free_user_key"
    )
    response = gateway.route_request(request)
    print(f"Status: {response.status_code}")
    print(f"Error: {response.error}")
    
    # Summary
    print("\n" + "="*80)
    print("SUMMARY: API Gateway Pattern Best Practices")
    print("="*80)
    print("""
1. ROUTING:
   - Centralized request routing
   - Service discovery and selection
   - Load balancing across backends
   - Flexible routing rules

2. AUTHENTICATION & AUTHORIZATION:
   - API key management
   - Multiple authorization levels
   - Fine-grained access control
   - Secure credential handling

3. RATE LIMITING:
   - Per-user rate limits
   - Per-route limits
   - Sliding window implementation
   - Graceful limit enforcement

4. MONITORING:
   - Request logging
   - Performance metrics
   - Service health tracking
   - Real-time visibility

5. ERROR HANDLING:
   - Standardized error responses
   - Proper HTTP status codes
   - Detailed error messages
   - Graceful degradation

6. SERVICE MANAGEMENT:
   - Service registration
   - Health checks
   - Dynamic service discovery
   - Version management

Benefits:
✓ Centralized control and management
✓ Security enforcement at gateway level
✓ Simplified client integration
✓ Traffic management and optimization
✓ Monitoring and observability
✓ Service abstraction

Challenges:
- Single point of failure risk
- Added network latency
- Gateway complexity
- Potential bottleneck
- Requires high availability
    """)


if __name__ == "__main__":
    demonstrate_api_gateway()

Agentic Design Pattern: Api Gateway

Pattern 156: Api Gateway – Line-by-Line Explanation

Overview

This document provides a detailed line-by-line explanation of the Api Gateway pattern implementation.

Pattern Description

Pattern 156: API Gateway Pattern

Description: Centralized gateway for managing API requests to multiple agents/services. Provides routing, authentication, rate limiting, monitoring, and unified interface for diverse backend services.

Components: – Request router – Authentication & authorization – Rate limiting – Load balancing – Monitoring & logging – Response transformation – Error handling

Use Cases: – Multi-agent systems – Microservices architecture – API versioning – Security enforcement – Traffic management

Benefits: – Centralized control – Security enforcement – Traffic management – Simplified client interface – Monitoring visibility

Trade-offs: – Single point of failure – Added latency – Complexity overhead – Gateway bottleneck potential

LangChain Implementation: Uses routing logic, rate limiting, authentication checks, and multiple LLM backends with unified interface

File Statistics

  • Total Lines: 632
  • Classes: 8
  • Functions: 1
  • Constants: 0

Module Documentation

"""
Pattern 156: API Gateway Pattern

Description:
    Centralized gateway for managing API requests to multiple agents/services.
    Provides routing, authentication, rate limiting, monitoring, and unified
    interface for diverse backend services.

Components:
    - Request router
    - Authentication & authorization
    - Rate limiting
    - Load balancing
    - Monitoring & logging
    - Response transformation
    - Error handling

Use Cases:
    - Multi-agent systems
    - Microservices architecture
    - API versioning
    - Security enforcement
    - Traffic management

Benefits:
    - Centralized control
    - Security enforcement
    - Traffic management
    - Simplified client interface
    - Monitoring visibility

Trade-offs:
    - Single point of failure
    - Added latency
    - Complexity overhead
    - Gateway bottleneck potential

LangChain Implementation:
    Uses routing logic, rate limiting, authentication checks,
    and multiple LLM backends with unified interface
"""

Purpose: This module implements the Api Gateway pattern, providing a structured approach to agent design and implementation.

Import Statements

Standard Library Imports

import collections.defaultdict
import dataclasses.dataclass
import dataclasses.field
import datetime.datetime
import datetime.timedelta
import enum.Enum
import langchain_core.output_parsers.StrOutputParser
import langchain_core.prompts.ChatPromptTemplate
import os
import random
import time
import typing.Any
import typing.Callable
import typing.Dict
import typing.List
import typing.Optional

Purpose: Core Python functionality: – collections.defaultdict: Specialized container datatypes – dataclasses.dataclass: Simplified class definitions with @dataclass – dataclasses.field: Simplified class definitions with @dataclass – datetime.datetime: Date and time handling – datetime.timedelta: Date and time handling – enum.Enum: Enumeration support – langchain_core.output_parsers.StrOutputParser: Regular expression operations – langchain_core.prompts.ChatPromptTemplate: Regular expression operations – os: Environment variable access and system operations – random: Random number generation – time: Time-related functions – typing.Any: Type hints for better code documentation (List, Dict, Any, Optional) – typing.Callable: Type hints for better code documentation (List, Dict, Any, Optional) – typing.Dict: Type hints for better code documentation (List, Dict, Any, Optional) – typing.List: Type hints for better code documentation (List, Dict, Any, Optional) – typing.Optional: Type hints for better code documentation (List, Dict, Any, Optional)

LangChain Imports

from langchain_openai.ChatOpenAI

Purpose: LangChain framework components: – LLM Interface: Wrapper for OpenAI models – Chains: Composable building blocks for complex workflows

Other Imports

from dotenv.load_dotenv

Purpose: Additional dependencies for specialized functionality.

Classes

Class: ServiceType

Lines 59-65

Inherits from: Enum

Purpose:

Types of backend services

Class: AuthLevel

Lines 68-73

Inherits from: Enum

Purpose:

Authentication levels

Class: Route

Lines 77-84

Purpose:

API route configuration

Class: ServiceConfig

Lines 88-95

Purpose:

Backend service configuration

Class: Request

Lines 99-107

Purpose:

Gateway request

Class: Response

Lines 111-118

Purpose:

Gateway response

Class: RateLimitInfo

Lines 122-126

Purpose:

Rate limit tracking

Class: APIGateway

Lines 129-452

Purpose:

Centralized API gateway for agent services

Methods:

__init__(self)

Lines 132-146

Purpose:

Initialize API gateway

Parameters:

_initialize_services(self)

Lines 148-172

Purpose:

Initialize backend services

Parameters:

_setup_routes(self)

Lines 174-205

Purpose:

Set up API routes

Parameters:

register_api_key(self, api_key, auth_level)

Lines 207-209

Purpose:

Register an API key with authentication level

Parameters: – api_key: Api key – auth_level: Auth level

authenticate(self, request)

Lines 211-240

Purpose:

Authenticate request

Parameters: – request: Request

Returns: <ast.Subscript object at 0x000001A06D65AE50>

check_rate_limit(self, request)

Lines 242-270

Purpose:

Check if request exceeds rate limit

Parameters: – request: Request

Returns: <ast.Subscript object at 0x000001A06D7CD210>

route_request(self, request)

Lines 272-339

Purpose:

Route and process request through gateway

Parameters: – request: Request

Returns: Response

_select_service(self, service_type)

Lines 341-355

Purpose:

Select service based on type and load balancing

Parameters: – service_type: Service type

Returns: <ast.Subscript object at 0x000001A06D6FDB90>

_handle_chat(self, request, service)

Lines 357-367

Purpose:

Handle chat request

Parameters: – request: Request – service: Service

Returns: str

_handle_analysis(self, request, service)

Lines 369-379

Purpose:

Handle analysis request

Parameters: – request: Request – service: Service

Returns: str

_handle_translation(self, request, service)

Lines 381-395

Purpose:

Handle translation request

Parameters: – request: Request – service: Service

Returns: str

_handle_summarization(self, request, service)

Lines 397-410

Purpose:

Handle summarization request

Parameters: – request: Request – service: Service

Returns: str

_log_request(self, request)

Lines 412-420

Purpose:

Log request for monitoring

Parameters: – request: Request

get_metrics(self)

Lines 422-452

Purpose:

Get gateway metrics

Parameters:

Returns: <ast.Subscript object at 0x000001A06D60D450>

Functions

demonstrate_api_gateway()

Lines 455-627

Purpose:

Demonstrate API gateway pattern

Key Concepts Summary

Pattern Overview

The Api Gateway pattern provides a structured approach to building AI agents with specific capabilities and behaviors. This implementation leverages LangChain’s composable components to create a robust and maintainable solution.

Core Components

  1. Agent Class: Main implementation of the pattern logic
  2. Prompt Engineering: Carefully crafted prompts to guide LLM behavior
  3. Tool Integration: Optional tools for extended functionality
  4. Memory Management: State handling across interactions
  5. Output Parsing: Structured response processing

When to Use

✅ Good for: – Scenarios requiring this specific pattern’s strengths – Applications needing structured agent behavior – Systems where explainability is important – Production environments requiring reliability

❌ Not ideal for: – Simple single-prompt tasks – Scenarios requiring different pattern characteristics – Resource-constrained environments (if computationally intensive)

Best Practices Demonstrated

1. Environment Configuration

  • ✅ Use .env files for API keys
  • ✅ Never hardcode credentials
  • ✅ Validate environment variables on startup

2. Error Handling

  • ✅ Graceful degradation on failures
  • ✅ Informative error messages
  • ✅ Logging for debugging

3. Prompt Engineering

  • ✅ Clear, specific instructions
  • ✅ Structured output formats
  • ✅ Few-shot examples when beneficial

4. Code Organization

  • ✅ Separation of concerns (class-based design)
  • ✅ Type hints for clarity
  • ✅ Docstrings for documentation

5. Performance Optimization

  • ✅ Appropriate temperature settings
  • ✅ Token budget management
  • ✅ Caching where applicable

LangChain Components Used

  • ChatOpenAI: OpenAI model interface
  • Chains: Sequential component composition

Why LangChain? – Composable abstractions for LLM applications – Rich ecosystem of integrations – Production-ready patterns – Active community and development

Execution Flow

1. Initialize agent with configuration
   ↓
2. Load environment and dependencies
   ↓
3. Prepare prompt with input data
   ↓
4. Invoke LLM through LangChain
   ↓
5. Parse and validate response
   ↓
6. Return formatted result

Key Decision Points: – Model selection (temperature, model type) – Prompt structure (few-shot vs zero-shot) – Error handling strategy – Output format validation

Performance Considerations

Computational Cost

  • LLM Calls: Primary cost factor (based on tokens)
  • Latency: Depends on model and input size
  • Memory: Minimal for most patterns

Optimization Strategies

  1. Prompt Optimization: Minimize tokens while maintaining clarity
  2. Caching: Cache repeated queries/responses
  3. Model Selection: Use smaller models when appropriate
  4. Batch Processing: Group similar requests
  5. Streaming: For long responses, stream tokens

Cost Estimates (Approximate)

  • GPT-3.5-turbo: ~$0.002 per request
  • GPT-4: ~$0.03 per request
  • GPT-4-turbo: ~$0.01 per request

Actual costs vary based on input/output token counts

Common Pitfalls to Avoid

1. Prompt Design Issues

❌ Vague or ambiguous instructions
✅ Clear, specific, testable prompts

2. Error Handling

❌ Ignoring API errors or rate limits
✅ Implement retry logic and graceful degradation

3. Token Management

❌ Exceeding model token limits
✅ Truncate inputs, chunk processing

4. Security

❌ Exposing API keys in code
✅ Use environment variables

5. Testing

❌ Only testing happy paths
✅ Test edge cases, failures, malformed inputs

6. Cost Management

❌ Unmonitored API usage
✅ Implement budgets, alerts, and monitoring

Further Improvements

Production Enhancements

  1. Monitoring: Add logging, metrics, tracing
  2. Validation: Input/output schema validation
  3. Rate Limiting: Implement request throttling
  4. Caching: Redis or in-memory cache for responses
  5. Async Support: Non-blocking operations for scalability

Feature Extensions

  1. Multi-Model Support: Switch between providers
  2. Fallback Models: Graceful degradation strategy
  3. Custom Tools: Domain-specific function calling
  4. Advanced Memory: Vector stores for long-term memory
  5. Evaluation: Automated testing and benchmarking

Code Quality

  1. Testing: Unit tests, integration tests
  2. Documentation: API docs, usage examples
  3. Type Safety: Stricter type hints
  4. Configuration: Externalized config files

Related Patterns

  • Pattern 155
  • Pattern 157

Pattern Families: – Reasoning Patterns (001-011) – Multi-Agent Patterns (012-019) – Tool Use Patterns (020-025) – Memory Patterns (026-031) – And many more…

Summary

The Api Gateway pattern provides a powerful approach to building AI agents with specific capabilities. This implementation demonstrates:

✅ Key Strengths: – Clean, maintainable code structure – LangChain integration for composability – Type hints and documentation – Production-ready patterns

🎯 Best Use Cases: – Applications requiring this pattern’s characteristics – Systems needing structured agent behavior – Production environments requiring reliability

📚 Learning Resources: – LangChain Documentation: https://python.langchain.com/ – Pattern Documentation: See related patterns – OpenAI API Reference: https://platform.openai.com/docs/

File: 156_api_gateway.py
Total Lines: 632
Pattern Category: AI Agent Design Pattern #156

Generated from actual implementation code
For educational and reference purposes

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *

More posts