"""
Pattern 079: Batch Processing Agent
Description:
A Batch Processing Agent handles multiple tasks or inputs concurrently in batches,
optimizing throughput and resource utilization. This pattern is essential for
processing large volumes of data efficiently, enabling parallel execution,
job queuing, progress monitoring, and result aggregation.
The agent manages batches of work items, distributes them across available resources,
monitors progress, handles failures, and aggregates results. It supports various
batching strategies including size-based, time-based, and priority-based batching.
Components:
1. Batch Scheduler: Schedules and organizes batch jobs
2. Job Queue: Manages pending work items
3. Parallel Executor: Executes tasks concurrently
4. Progress Monitor: Tracks batch processing progress
5. Result Aggregator: Combines individual results
6. Error Handler: Handles failures and retries
7. Resource Manager: Manages compute resources
8. Throughput Optimizer: Optimizes batch sizes
Key Features:
- Parallel batch processing
- Dynamic batch sizing
- Job queue management
- Progress tracking
- Error handling and retries
- Result aggregation
- Resource optimization
- Priority scheduling
- Cancellation support
- Checkpointing for resumption
Use Cases:
- Bulk document processing
- Large-scale data analysis
- Batch translation services
- Bulk content generation
- Mass email personalization
- Dataset annotation
- Batch image processing
- ETL pipelines
- Bulk API operations
- Large-scale testing
LangChain Implementation:
Uses concurrent.futures for parallel execution, batch processing utilities,
and structured output handling for aggregation.
"""
import os
import time
import uuid
from typing import List, Dict, Any, Optional, Callable, Tuple
from dataclasses import dataclass, field
from enum import Enum
from datetime import datetime
from concurrent.futures import ThreadPoolExecutor, as_completed
from collections import defaultdict
from dotenv import load_dotenv
from langchain_core.prompts import ChatPromptTemplate
from langchain_core.output_parsers import StrOutputParser
from langchain_openai import ChatOpenAI
load_dotenv()
class JobStatus(Enum):
"""Status of a batch job"""
PENDING = "pending"
QUEUED = "queued"
PROCESSING = "processing"
COMPLETED = "completed"
FAILED = "failed"
CANCELLED = "cancelled"
RETRYING = "retrying"
class BatchStrategy(Enum):
"""Batching strategy"""
SIZE_BASED = "size_based" # Fixed batch size
TIME_BASED = "time_based" # Time window
ADAPTIVE = "adaptive" # Dynamic sizing
PRIORITY = "priority" # Priority-based
@dataclass
class BatchJob:
"""A single job in a batch"""
job_id: str
input_data: Any
priority: int = 0 # Higher = more important
status: JobStatus = JobStatus.PENDING
result: Optional[Any] = None
error: Optional[str] = None
attempts: int = 0
created_at: datetime = field(default_factory=datetime.now)
started_at: Optional[datetime] = None
completed_at: Optional[datetime] = None
@dataclass
class BatchConfig:
"""Configuration for batch processing"""
batch_size: int = 10
max_workers: int = 4
max_retries: int = 3
timeout: float = 60.0 # seconds per job
strategy: BatchStrategy = BatchStrategy.SIZE_BASED
aggregate_results: bool = True
checkpoint_interval: int = 10 # Jobs between checkpoints
@dataclass
class BatchResult:
"""Result of batch processing"""
batch_id: str
total_jobs: int
completed: int
failed: int
cancelled: int
results: List[Any]
errors: List[Tuple[str, str]] # (job_id, error)
duration: float
throughput: float # jobs/second
metadata: Dict[str, Any] = field(default_factory=dict)
@dataclass
class BatchProgress:
"""Progress of batch processing"""
batch_id: str
total_jobs: int
completed: int
processing: int
pending: int
failed: int
progress_pct: float
estimated_remaining: float # seconds
current_throughput: float
class BatchProcessingAgent:
"""
Agent for efficient batch processing of multiple tasks.
This agent handles large volumes of work by processing items in batches,
with parallel execution, error handling, and progress tracking.
"""
def __init__(self, config: BatchConfig = BatchConfig()):
"""Initialize the batch processing agent"""
self.config = config
# LLM for processing
self.llm = ChatOpenAI(
model="gpt-3.5-turbo",
temperature=0.3
)
# Job management
self.jobs: Dict[str, BatchJob] = {}
self.batches: Dict[str, List[str]] = {} # batch_id -> job_ids
self.active_batches: List[str] = []
def create_batch(
self,
inputs: List[Any],
priorities: Optional[List[int]] = None
) -> str:
"""
Create a new batch of jobs.
Args:
inputs: List of input data for jobs
priorities: Optional priority for each job
Returns:
Batch ID
"""
batch_id = f"batch_{uuid.uuid4().hex[:8]}"
job_ids = []
for i, input_data in enumerate(inputs):
job_id = f"job_{uuid.uuid4().hex[:8]}"
priority = priorities[i] if priorities else 0
job = BatchJob(
job_id=job_id,
input_data=input_data,
priority=priority,
status=JobStatus.QUEUED
)
self.jobs[job_id] = job
job_ids.append(job_id)
self.batches[batch_id] = job_ids
self.active_batches.append(batch_id)
return batch_id
def process_batch(
self,
batch_id: str,
task_fn: Callable[[Any], Any],
on_progress: Optional[Callable[[BatchProgress], None]] = None
) -> BatchResult:
"""
Process a batch of jobs in parallel.
Args:
batch_id: ID of batch to process
task_fn: Function to process each job
on_progress: Optional progress callback
Returns:
BatchResult with aggregated results
"""
if batch_id not in self.batches:
raise ValueError(f"Batch {batch_id} not found")
job_ids = self.batches[batch_id]
start_time = time.time()
results = []
errors = []
completed_count = 0
failed_count = 0
cancelled_count = 0
# Process jobs in parallel
with ThreadPoolExecutor(max_workers=self.config.max_workers) as executor:
# Submit jobs
future_to_job = {}
for job_id in job_ids:
job = self.jobs[job_id]
job.status = JobStatus.PROCESSING
job.started_at = datetime.now()
future = executor.submit(
self._process_single_job,
job_id,
task_fn
)
future_to_job[future] = job_id
# Collect results as they complete
for future in as_completed(future_to_job):
job_id = future_to_job[future]
job = self.jobs[job_id]
try:
result = future.result(timeout=self.config.timeout)
job.result = result
job.status = JobStatus.COMPLETED
job.completed_at = datetime.now()
results.append(result)
completed_count += 1
except Exception as e:
job.error = str(e)
job.status = JobStatus.FAILED
job.completed_at = datetime.now()
errors.append((job_id, str(e)))
failed_count += 1
# Progress callback
if on_progress:
progress = self._calculate_progress(
batch_id,
start_time,
completed_count + failed_count
)
on_progress(progress)
duration = time.time() - start_time
total_jobs = len(job_ids)
return BatchResult(
batch_id=batch_id,
total_jobs=total_jobs,
completed=completed_count,
failed=failed_count,
cancelled=cancelled_count,
results=results,
errors=errors,
duration=duration,
throughput=total_jobs / duration if duration > 0 else 0
)
def _process_single_job(
self,
job_id: str,
task_fn: Callable[[Any], Any]
) -> Any:
"""
Process a single job with retries.
Args:
job_id: Job ID
task_fn: Task function
Returns:
Job result
"""
job = self.jobs[job_id]
for attempt in range(self.config.max_retries):
try:
job.attempts = attempt + 1
if attempt > 0:
job.status = JobStatus.RETRYING
result = task_fn(job.input_data)
return result
except Exception as e:
if attempt == self.config.max_retries - 1:
raise
time.sleep(2 ** attempt) # Exponential backoff
raise RuntimeError(f"Job {job_id} failed after {self.config.max_retries} attempts")
def _calculate_progress(
self,
batch_id: str,
start_time: float,
completed: int
) -> BatchProgress:
"""Calculate batch progress"""
job_ids = self.batches[batch_id]
total = len(job_ids)
processing = sum(
1 for jid in job_ids
if self.jobs[jid].status == JobStatus.PROCESSING
)
pending = sum(
1 for jid in job_ids
if self.jobs[jid].status in [JobStatus.PENDING, JobStatus.QUEUED]
)
failed = sum(
1 for jid in job_ids
if self.jobs[jid].status == JobStatus.FAILED
)
progress_pct = (completed / total) * 100 if total > 0 else 0
elapsed = time.time() - start_time
throughput = completed / elapsed if elapsed > 0 else 0
remaining = total - completed
estimated_remaining = remaining / throughput if throughput > 0 else 0
return BatchProgress(
batch_id=batch_id,
total_jobs=total,
completed=completed,
processing=processing,
pending=pending,
failed=failed,
progress_pct=progress_pct,
estimated_remaining=estimated_remaining,
current_throughput=throughput
)
def process_text_batch(
self,
texts: List[str],
operation: str,
on_progress: Optional[Callable[[BatchProgress], None]] = None
) -> BatchResult:
"""
Process a batch of text operations.
Args:
texts: List of text inputs
operation: Operation to perform (summarize, translate, etc.)
on_progress: Optional progress callback
Returns:
BatchResult with processed texts
"""
# Create batch
batch_id = self.create_batch(texts)
# Define task function
def process_text(text: str) -> str:
prompt = ChatPromptTemplate.from_template(
"{operation} the following text:\n\n{text}"
)
chain = prompt | self.llm | StrOutputParser()
result = chain.invoke({
"operation": operation,
"text": text[:500] # Limit input length
})
return result
# Process batch
return self.process_batch(batch_id, process_text, on_progress)
def process_classification_batch(
self,
items: List[str],
categories: List[str],
on_progress: Optional[Callable[[BatchProgress], None]] = None
) -> BatchResult:
"""
Batch classify items into categories.
Args:
items: Items to classify
categories: Possible categories
on_progress: Optional progress callback
Returns:
BatchResult with classifications
"""
batch_id = self.create_batch(items)
categories_str = ", ".join(categories)
def classify_item(item: str) -> str:
prompt = ChatPromptTemplate.from_template(
"""Classify the following item into one of these categories: {categories}
Item: {item}
Category:"""
)
chain = prompt | self.llm | StrOutputParser()
result = chain.invoke({
"categories": categories_str,
"item": item[:200]
})
return result.strip()
return self.process_batch(batch_id, classify_item, on_progress)
def process_qa_batch(
self,
questions: List[str],
context: str,
on_progress: Optional[Callable[[BatchProgress], None]] = None
) -> BatchResult:
"""
Answer a batch of questions.
Args:
questions: List of questions
context: Shared context for all questions
on_progress: Optional progress callback
Returns:
BatchResult with answers
"""
batch_id = self.create_batch(questions)
def answer_question(question: str) -> str:
prompt = ChatPromptTemplate.from_template(
"""Context: {context}
Question: {question}
Answer:"""
)
chain = prompt | self.llm | StrOutputParser()
result = chain.invoke({
"context": context[:1000],
"question": question
})
return result
return self.process_batch(batch_id, answer_question, on_progress)
def get_batch_status(self, batch_id: str) -> Dict[str, Any]:
"""
Get status of a batch.
Args:
batch_id: Batch ID
Returns:
Status dictionary
"""
if batch_id not in self.batches:
raise ValueError(f"Batch {batch_id} not found")
job_ids = self.batches[batch_id]
status_counts = defaultdict(int)
for job_id in job_ids:
status = self.jobs[job_id].status
status_counts[status.value] += 1
return {
"batch_id": batch_id,
"total_jobs": len(job_ids),
"status_breakdown": dict(status_counts),
"jobs": [
{
"job_id": jid,
"status": self.jobs[jid].status.value,
"attempts": self.jobs[jid].attempts,
"error": self.jobs[jid].error
}
for jid in job_ids[:10] # First 10 jobs
]
}
def demonstrate_batch_processing_agent():
"""Demonstrate the batch processing agent capabilities"""
print("=" * 80)
print("BATCH PROCESSING AGENT DEMONSTRATION")
print("=" * 80)
# Custom config for faster demos
config = BatchConfig(
batch_size=5,
max_workers=3,
max_retries=2
)
agent = BatchProcessingAgent(config)
# Demo 1: Text Summarization Batch
print("\n" + "=" * 80)
print("DEMO 1: Batch Text Summarization")
print("=" * 80)
texts = [
"Artificial intelligence is transforming healthcare through advanced diagnostic tools, personalized treatment plans, and drug discovery acceleration.",
"Machine learning models require large amounts of training data and computational resources to achieve high performance on complex tasks.",
"Natural language processing enables computers to understand, interpret, and generate human language in meaningful ways.",
"Computer vision systems can now recognize objects, faces, and scenes with accuracy rivaling human perception.",
"Robotics combined with AI is enabling autonomous vehicles, warehouse automation, and advanced manufacturing processes."
]
print(f"\nProcessing {len(texts)} texts for summarization...")
print("-" * 80)
progress_updates = []
def on_progress(progress: BatchProgress):
progress_updates.append(progress)
bar_length = 40
filled = int(bar_length * progress.progress_pct / 100)
bar = "█" * filled + "░" * (bar_length - filled)
print(f"\r[{bar}] {progress.progress_pct:.1f}% ({progress.completed}/{progress.total_jobs})", end="", flush=True)
result = agent.process_text_batch(texts, "Summarize in one sentence", on_progress)
print("\n" + "-" * 80)
print(f"\n✓ Batch complete!")
print(f"Total Jobs: {result.total_jobs}")
print(f"Completed: {result.completed}")
print(f"Failed: {result.failed}")
print(f"Duration: {result.duration:.2f}s")
print(f"Throughput: {result.throughput:.2f} jobs/sec")
print(f"\nSample Results:")
for i, summary in enumerate(result.results[:3], 1):
print(f"\n{i}. {summary[:100]}...")
# Demo 2: Classification Batch
print("\n" + "=" * 80)
print("DEMO 2: Batch Classification")
print("=" * 80)
items = [
"Python programming tutorial",
"Chocolate chip cookie recipe",
"2024 Olympic games highlights",
"Climate change research paper",
"Stock market analysis",
"Travel guide to Japan",
"Guitar lesson for beginners"
]
categories = ["Technology", "Food", "Sports", "Science", "Finance", "Travel", "Music"]
print(f"\nClassifying {len(items)} items into categories:")
print(f"Categories: {', '.join(categories)}")
print("-" * 80)
result2 = agent.process_classification_batch(items, categories, on_progress=on_progress)
print("\n" + "-" * 80)
print(f"\n✓ Classification complete!")
print(f"Duration: {result2.duration:.2f}s")
print(f"Throughput: {result2.throughput:.2f} items/sec")
print(f"\nClassification Results:")
for i, (item, category) in enumerate(zip(items, result2.results), 1):
print(f"{i}. '{item[:40]}...' → {category}")
# Demo 3: Question Answering Batch
print("\n" + "=" * 80)
print("DEMO 3: Batch Question Answering")
print("=" * 80)
context = """
LangChain is a framework for developing applications powered by language models.
It provides tools for prompt management, chains, agents, memory, and more.
LangChain supports multiple LLM providers including OpenAI, Anthropic, and Google.
The framework emphasizes composability and modularity for building complex applications.
"""
questions = [
"What is LangChain?",
"What providers does LangChain support?",
"What are the main features of LangChain?",
"What is the design philosophy of LangChain?"
]
print(f"\nContext: {context.strip()[:100]}...")
print(f"\nAnswering {len(questions)} questions...")
print("-" * 80)
result3 = agent.process_qa_batch(questions, context, on_progress=on_progress)
print("\n" + "-" * 80)
print(f"\n✓ Q&A batch complete!")
print(f"Duration: {result3.duration:.2f}s")
print(f"\nQuestions and Answers:")
for i, (question, answer) in enumerate(zip(questions, result3.results), 1):
print(f"\nQ{i}: {question}")
print(f"A{i}: {answer[:150]}...")
# Demo 4: Batch Status Monitoring
print("\n" + "=" * 80)
print("DEMO 4: Batch Status Monitoring")
print("=" * 80)
# Create a new batch
test_items = ["Item " + str(i) for i in range(8)]
batch_id = agent.create_batch(test_items)
print(f"\nCreated batch: {batch_id}")
print(f"Total items: {len(test_items)}")
# Get status before processing
status = agent.get_batch_status(batch_id)
print(f"\nBatch Status:")
print(f" Total Jobs: {status['total_jobs']}")
print(f" Status Breakdown:")
for status_type, count in status['status_breakdown'].items():
print(f" {status_type}: {count}")
# Demo 5: Performance Comparison
print("\n" + "=" * 80)
print("DEMO 5: Parallel vs Sequential Performance")
print("=" * 80)
test_texts = [f"Sample text {i} for processing" for i in range(10)]
print(f"\nProcessing {len(test_texts)} items...")
# Parallel processing
print("\nParallel Processing (4 workers):")
parallel_start = time.time()
parallel_result = agent.process_text_batch(test_texts, "Count words in")
parallel_duration = time.time() - parallel_start
print(f" Duration: {parallel_duration:.2f}s")
print(f" Throughput: {parallel_result.throughput:.2f} jobs/sec")
# Sequential processing (1 worker)
print("\nSequential Processing (1 worker):")
sequential_config = BatchConfig(max_workers=1)
sequential_agent = BatchProcessingAgent(sequential_config)
sequential_start = time.time()
sequential_result = sequential_agent.process_text_batch(test_texts, "Count words in")
sequential_duration = time.time() - sequential_start
print(f" Duration: {sequential_duration:.2f}s")
print(f" Throughput: {sequential_result.throughput:.2f} jobs/sec")
speedup = sequential_duration / parallel_duration if parallel_duration > 0 else 0
print(f"\nSpeedup: {speedup:.2f}x")
print(f"Efficiency: {(speedup / config.max_workers) * 100:.1f}%")
# Summary
print("\n" + "=" * 80)
print("SUMMARY")
print("=" * 80)
summary = """
The Batch Processing Agent demonstrates efficient parallel processing:
KEY CAPABILITIES:
1. Parallel Execution: Process multiple jobs concurrently
2. Job Queuing: Manage large volumes of pending work
3. Progress Tracking: Real-time monitoring with callbacks
4. Error Handling: Automatic retries with exponential backoff
5. Result Aggregation: Collect and organize batch results
6. Resource Management: Control worker threads and resources
7. Flexible Batching: Support various batching strategies
8. Status Monitoring: Track job and batch status
BENEFITS:
- Dramatically improved throughput
- Efficient resource utilization
- Scalable processing of large datasets
- Automatic error recovery
- Progress visibility
- Cost optimization through parallelization
- Reduced total processing time
- Better user experience for bulk operations
USE CASES:
- Bulk document summarization or translation
- Large-scale data classification
- Mass content generation
- Batch question answering
- Dataset annotation and labeling
- Bulk email personalization
- ETL pipeline processing
- Large-scale testing
- Batch API operations
- Document ingestion for RAG systems
PRODUCTION CONSIDERATIONS:
1. Worker Scaling: Adjust workers based on load and resources
2. Rate Limiting: Respect API rate limits across workers
3. Cost Management: Monitor token usage across batches
4. Error Recovery: Implement robust retry strategies
5. Checkpointing: Save progress for resumption
6. Priority Queues: Process high-priority jobs first
7. Resource Limits: Prevent resource exhaustion
8. Monitoring: Track batch performance metrics
9. Load Balancing: Distribute work evenly
10. Graceful Shutdown: Handle interruptions cleanly
ADVANCED EXTENSIONS:
- Dynamic worker scaling based on queue depth
- Intelligent batch sizing with machine learning
- Multi-stage batch pipelines
- Distributed processing across machines
- Streaming batch results
- Partial result delivery
- Batch deduplication
- Priority-based scheduling
- Resource-aware batching
- Automatic retry backoff tuning
Batch processing is essential for production AI systems handling
large volumes of data efficiently and cost-effectively.
"""
print(summary)
if __name__ == "__main__":
demonstrate_batch_processing_agent()
Agentic Design Pattern: Batch Processing
Pattern 079: Batch Processing – Line-by-Line Explanation
Overview
This document provides a detailed line-by-line explanation of the Batch Processing pattern implementation.
Pattern Description
Pattern 079: Batch Processing Agent
Description: A Batch Processing Agent handles multiple tasks or inputs concurrently in batches, optimizing throughput and resource utilization. This pattern is essential for processing large volumes of data efficiently, enabling parallel execution, job queuing, progress monitoring, and result aggregation.
The agent manages batches of work items, distributes them across available resources,
monitors progress, handles failures, and aggregates results. It supports various
batching strategies including size-based, time-based, and priority-based batching.
Components: 1. Batch Scheduler: Schedules and organizes batch jobs 2. Job Queue: Manages pending work items 3. Parallel Executor: Executes tasks concurrently 4. Progress Monitor: Tracks batch processing progress 5. Result Aggregator: Combines individual results 6. Error Handler: Handles failures and retries 7. Resource Manager: Manages compute resources 8. Throughput Optimizer: Optimizes batch sizes
Key Features: – Parallel batch processing – Dynamic batch sizing – Job queue management – Progress tracking – Error handling and retries – Result aggregation – Resource optimization – Priority scheduling – Cancellation support – Checkpointing for resumption
Use Cases: – Bulk document processing – Large-scale data analysis – Batch translation services – Bulk content generation – Mass email personalization – Dataset annotation – Batch image processing – ETL pipelines – Bulk API operations – Large-scale testing
LangChain Implementation: Uses concurrent.futures for parallel execution, batch processing utilities, and structured output handling for aggregation.
File Statistics
- Total Lines: 773
- Classes: 7
- Functions: 1
- Constants: 0
Module Documentation
"""
Pattern 079: Batch Processing Agent
Description:
A Batch Processing Agent handles multiple tasks or inputs concurrently in batches,
optimizing throughput and resource utilization. This pattern is essential for
processing large volumes of data efficiently, enabling parallel execution,
job queuing, progress monitoring, and result aggregation.
The agent manages batches of work items, distributes them across available resources,
monitors progress, handles failures, and aggregates results. It supports various
batching strategies including size-based, time-based, and priority-based batching.
Components:
1. Batch Scheduler: Schedules and organizes batch jobs
2. Job Queue: Manages pending work items
3. Parallel Executor: Executes tasks concurrently
4. Progress Monitor: Tracks batch processing progress
5. Result Aggregator: Combines individual results
6. Error Handler: Handles failures and retries
7. Resource Manager: Manages compute resources
8. Throughput Optimizer: Optimizes batch sizes
Key Features:
- Parallel batch processing
- Dynamic batch sizing
- Job queue management
- Progress tracking
- Error handling and retries
- Result aggregation
- Resource optimization
- Priority scheduling
- Cancellation support
- Checkpointing for resumption
Use Cases:
- Bulk document processing
- Large-scale data analysis
- Batch translation services
- Bulk content generation
- Mass email personalization
- Dataset annotation
- Batch image processing
- ETL pipelines
- Bulk API operations
- Large-scale testing
LangChain Implementation:
Uses concurrent.futures for parallel execution, batch processing utilities,
and structured output handling for aggregation.
"""
Purpose: This module implements the Batch Processing pattern, providing a structured approach to agent design and implementation.
Import Statements
Standard Library Imports
import collections.defaultdict
import concurrent.futures.ThreadPoolExecutor
import concurrent.futures.as_completed
import dataclasses.dataclass
import dataclasses.field
import datetime.datetime
import enum.Enum
import langchain_core.output_parsers.StrOutputParser
import langchain_core.prompts.ChatPromptTemplate
import os
import time
import typing.Any
import typing.Callable
import typing.Dict
import typing.List
import typing.Optional
import typing.Tuple
Purpose: Core Python functionality:
– collections.defaultdict: Specialized container datatypes
– concurrent.futures.ThreadPoolExecutor: Regular expression operations
– concurrent.futures.as_completed: Regular expression operations
– dataclasses.dataclass: Simplified class definitions with @dataclass
– dataclasses.field: Simplified class definitions with @dataclass
– datetime.datetime: 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
– 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)
– typing.Tuple: 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
from uuid
Purpose: Additional dependencies for specialized functionality.
Classes
Class: JobStatus
Lines 71-79
Inherits from: Enum
Purpose:
Status of a batch job
Class: BatchStrategy
Lines 82-87
Inherits from: Enum
Purpose:
Batching strategy
Class: BatchJob
Lines 91-102
Purpose:
A single job in a batch
Class: BatchConfig
Lines 106-114
Purpose:
Configuration for batch processing
Class: BatchResult
Lines 118-129
Purpose:
Result of batch processing
Class: BatchProgress
Lines 133-143
Purpose:
Progress of batch processing
Class: BatchProcessingAgent
Lines 146-523
Purpose:
Agent for efficient batch processing of multiple tasks.
This agent handles large volumes of work by processing items in batches,
with parallel execution, error handling, and progress tracking.
Methods:
__init__(self, config)
Lines 154-167
Purpose:
Initialize the batch processing agent
Parameters:
– config: Config
create_batch(self, inputs, priorities)
Lines 169-204
Purpose:
Create a new batch of jobs.
Args:
inputs: List of input data for jobs
priorities: Optional priority for each job
Returns:
Batch ID
Parameters:
– inputs: Input data
– priorities: Priorities
Returns: str
process_batch(self, batch_id, task_fn, on_progress)
Lines 206-293
Purpose:
Process a batch of jobs in parallel.
Args:
batch_id: ID of batch to process
task_fn: Function to process each job
on_progress: Optional progress callback
Returns:
BatchResult with aggregated results
Parameters:
– batch_id: Batch id
– task_fn: Task fn
– on_progress: On progress
Returns: BatchResult
_process_single_job(self, job_id, task_fn)
Lines 295-326
Purpose:
Process a single job with retries.
Args:
job_id: Job ID
task_fn: Task function
Returns:
Job result
Parameters:
– job_id: Job id
– task_fn: Task fn
Returns: Any
_calculate_progress(self, batch_id, start_time, completed)
Lines 328-370
Purpose:
Calculate batch progress
Parameters:
– batch_id: Batch id
– start_time: Start time
– completed: Completed
Returns: BatchProgress
process_text_batch(self, texts, operation, on_progress)
Lines 372-408
Purpose:
Process a batch of text operations.
Args:
texts: List of text inputs
operation: Operation to perform (summarize, translate, etc.)
on_progress: Optional progress callback
Returns:
BatchResult with processed texts
Parameters:
– texts: Texts
– operation: Operation
– on_progress: On progress
Returns: BatchResult
process_classification_batch(self, items, categories, on_progress)
Lines 410-449
Purpose:
Batch classify items into categories.
Args:
items: Items to classify
categories: Possible categories
on_progress: Optional progress callback
Returns:
BatchResult with classifications
Parameters:
– items: Items
– categories: Categories
– on_progress: On progress
Returns: BatchResult
process_qa_batch(self, questions, context, on_progress)
Lines 451-488
Purpose:
Answer a batch of questions.
Args:
questions: List of questions
context: Shared context for all questions
on_progress: Optional progress callback
Returns:
BatchResult with answers
Parameters:
– questions: Question to answer
– context: Context
– on_progress: On progress
Returns: BatchResult
get_batch_status(self, batch_id)
Lines 490-523
Purpose:
Get status of a batch.
Args:
batch_id: Batch ID
Returns:
Status dictionary
Parameters:
– batch_id: Batch id
Returns: <ast.Subscript object at 0x000001A06D62A990>
Functions
demonstrate_batch_processing_agent()
Lines 526-768
Purpose:
Demonstrate the batch processing agent capabilities
Key Concepts Summary
Pattern Overview
The Batch Processing 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
- Agent Class: Main implementation of the pattern logic
- Prompt Engineering: Carefully crafted prompts to guide LLM behavior
- Tool Integration: Optional tools for extended functionality
- Memory Management: State handling across interactions
- 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
.envfiles 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
- Prompt Optimization: Minimize tokens while maintaining clarity
- Caching: Cache repeated queries/responses
- Model Selection: Use smaller models when appropriate
- Batch Processing: Group similar requests
- 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
- Monitoring: Add logging, metrics, tracing
- Validation: Input/output schema validation
- Rate Limiting: Implement request throttling
- Caching: Redis or in-memory cache for responses
- Async Support: Non-blocking operations for scalability
Feature Extensions
- Multi-Model Support: Switch between providers
- Fallback Models: Graceful degradation strategy
- Custom Tools: Domain-specific function calling
- Advanced Memory: Vector stores for long-term memory
- Evaluation: Automated testing and benchmarking
Code Quality
- Testing: Unit tests, integration tests
- Documentation: API docs, usage examples
- Type Safety: Stricter type hints
- Configuration: Externalized config files
Related Patterns
- Pattern 078
- Pattern 080
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 Batch Processing 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: 079_batch_processing.py
Total Lines: 773
Pattern Category: AI Agent Design Pattern #079
Generated from actual implementation code
For educational and reference purposes
Leave a Reply