ai-foundation-kit 0.1.3

Common utilities for AI projects including logging and exceptions

🏗️ AI Foundation Kit

The Professional / Development-First Standard for AI/ML Applications

ai-foundation-kit is an enterprise-grade Python library designed to provide the bedrock for building scalable, robust, and maintainable AI systems. It unifies core utilities—logging, exception handling, file management, model loading, and configuration—under a single, cohesive namespace: AIFoundationKit.

Designed for: RAG, Generative AI, Agentic Workflows, Deep Learning, and Machine Learning.

---

📦 Installation

From PyPI (Recommended)

pip install ai-foundation-kit

From Source (Development)

git clone https://github.com/aignishant/ai-foundation-kit.git
cd ai-foundation-kit
pip install -e .

---

🏛️ Package Architecture

The package is structured under the AIFoundationKit namespace. Each submodule targets a specific domain of AI application development.

Module	Import Path	Description	Key Components
Base	AIFoundationKit.base	The Core Foundation. Essential utilities required by every application.	BaseFileManager, ApiKeyManager, AppException, Logger
RAG	AIFoundationKit.rag	Retrieval-Augmented Generation. Tools for LLM loading and prompting.	ModelLoader, BaseProvider, GenericPrompts
GenAI	AIFoundationKit.genai	Generative AI. (Upcoming) Specialized tools for image/video generation.	Planned
Agentic	AIFoundationKit.agentic	Agentic AI. (Upcoming) Frameworks for autonomous agents.	Planned
ML	AIFoundationKit.ml	Machine Learning. (Upcoming) Scikit-learn wrappers & utilities.	Planned
DL	AIFoundationKit.dl	Deep Learning. (Upcoming) Torch/Tensorflow helpers.	Planned

---

🛠️ Usage Guide

1. File Management (AIFoundationKit.base)

Stop writing boilerplate file readers. Use the BaseFileManager to handle PDFs, DOCX, JSON, and more seamlessly.

from AIFoundationKit.base.file_manager import BaseFileManager

file_manager = BaseFileManager()

# Read ANY file type (PDF, DOCX, CSV, TXT, etc.)
content = file_manager.read_file("path/to/document.pdf")
print(content[:100])

# Save files (handling bytes or file-like objects)
saved_path = file_manager.save_file(
    file_obj=b"Binary data", 
    save_dir="data/uploads", 
    file_name="invoice.pdf"
)

2. Model Loading (AIFoundationKit.rag)

Instantly switch between Google Gemini, Groq, or your own custom providers without changing application logic.

from AIFoundationKit.rag.model_loader import ModelLoader

# Initialize (loads config.yaml automatically or uses defaults)
loader = ModelLoader()

# Load LLM (Gemini Pro by default or configured)
llm = loader.load_llm(model_name="gemini-1.5-pro")

# Load Embeddings
embeddings = loader.load_embeddings()

# Use in LangChain
chain = prompt | llm

3. API Key Management (AIFoundationKit.base)

Securely manage keys from Environment Variables or Secret Stores.

from AIFoundationKit.base.model import ApiKeyManager

key_mgr = ApiKeyManager()
google_key = key_mgr.get("GOOGLE_API_KEY")

4. Enterprise Logging & Exceptions

Standardized structured logging and error handling.

from AIFoundationKit.base.logger.custom_logger import logger
from AIFoundationKit.base.exception.custom_exception import AppException

try:
    logger.info("Starting process", extra={"user_id": "12345"})
    # ... logic ...
except Exception as e:
    # Auto-wraps with standard error codes
    raise AppException(f"Process failed: {str(e)}") from e

---

🔮 Futuristic Example: The Agentic Workflow

Here is a glimpse of how you can build an autonomous file analysis agent using AIFoundationKit.

from AIFoundationKit.base.file_manager import BaseFileManager
from AIFoundationKit.rag.model_loader import ModelLoader
from AIFoundationKit.rag.prompts import get_generic_prompt

async def run_autonomous_analyst(file_path: str):
    # 1. Perception: Read the File
    fm = BaseFileManager()
    file_content = fm.read_file(file_path)
    
    # 2. Brain: Initialize Neural Core (LLM)
    loader = ModelLoader()
    llm = loader.load_llm(provider="groq", model_name="llama-3.3-70b-versatile")
    
    # 3. Cognition: Analyze with Generic Prompt
    analysis_prompt = get_generic_prompt("summary")
    chain = analysis_prompt | llm
    
    result = chain.invoke({"input_text": file_content})
    
    print("🤖 Analyst Report:")
    print(result.content)

# Execution
# await run_autonomous_analyst("contracts/q4_financials.pdf")

---

📝 Changelog

Version	Changes
v0.1.1	• Update: Config/Readme update (Stable). • Core: Includes base (File/Key Utils), rag (Model Loading), logging, exceptions.
v0.1.0	• Initial Release: Fresh launch of ai-foundation-kit (Stable). • Core: Includes base (File/Key Utils), rag (Model Loading), logging, exceptions.

---

License

MIT License