Projects

Open-source tools and patented innovations for enterprise AI systems. All projects are freely available for commercial use.

Open Source Projects

TOON Converter (Python)
Token Optimized Object Notation - A Python library that reduces LLM token consumption by 40-60% when sending structured data to AI systems.
PyPI

TOON (Token Optimized Object Notation) is a data format designed specifically for LLM communication. It separates schema from data, eliminating the repeated attribute names that make JSON verbose. The Python library handles nested objects, arrays, null values, and special characters automatically.

Install via PyPI:

pip install toon-token-optimizer

Key Features

  • 40-60% token reduction for structured data
  • Handles nested objects and arrays
  • 75+ unit tests for reliability
  • MIT license - free for commercial use
  • Full documentation and examples

Quick Example

from toon_converter import json_to_toon

customers = [
    {"name": "John", "age": 30, "city": "NYC"},
    {"name": "Jane", "age": 25, "city": "LA"},
]

toon_data = json_to_toon(customers)
# @schema:name,age,city
# John|30|NYC
# Jane|25|LA
PythonLLMToken OptimizationOpen Source
View on GitHubView on PyPI
TOON Converter (.NET)
The .NET/C# implementation of TOON Token Optimizer, bringing the same 40-60% token reduction to the .NET ecosystem.
NuGet

A C# implementation of the TOON format for .NET developers. Designed for enterprise .NET applications integrating with LLMs, this library provides the same token optimization benefits with idiomatic C# APIs and full .NET compatibility.

Install via NuGet:

dotnet add package Toon.TokenOptimizer

Key Features

  • Native .NET implementation
  • Token reduction statistics API
  • Supports anonymous types and POCOs
  • Comprehensive test coverage
  • Azure OpenAI integration examples

Quick Example

using Toon.TokenOptimizer;

var customers = new[] {
    new { Name = "John", Age = 30, City = "NYC" },
    new { Name = "Jane", Age = 25, City = "LA" }
};

string toonData = ToonConverter.ToToon(customers);

var stats = ToonConverter.GetTokenReduction(customers);
Console.WriteLine($"Saved {stats.ReductionPercent:F1}%");
C#.NETLLMToken OptimizationOpen Source
View on GitHubView on NuGet

Patented Innovation

U.S. Patent
Data Lake Management System
U.S. Patent No. 10,846,307 B1

This patent describes a novel approach to data lake architecture that addresses the 'garbage in, garbage out' problem that prevents enterprises from deploying reliable AI. The methodology has been cited by R&D teams at Fortune 500 companies including Intuit, Johnson & Johnson (Abiomed), and Mastercard.

Cited by Fortune 500 R&D Teams

Intuit Inc.

Financial Technology

US20230244661A1 (2023)

Abiomed / Johnson & Johnson

Medical Devices

DE112022000538T5 (2023)

Mastercard Technologies

Payments Infrastructure

US20240303248A1 (2024)

View Patent on Google Patents

Want to use these in your project?

All open-source projects are MIT licensed and free for commercial use. Check out the documentation or reach out if you need help with implementation.

Read the TOON GuideGet Implementation Help