The TQNN Core Engine is a protected trade secret. This repository provides only the API interface and client utilities. No inference substrate, internal algorithms, or architectural logic are present.
Unauthorized attempts to extract, replicate, or reverse-engineer substrate logic will result in immediate access termination and legal action.
🔹 TQNN AnyEngine API
Public SaaS API for TQNN AnyEngine Modes supported: EEG / Finance / Tabular / Image / Any-Data
🌐 Overview
TQNN — Tubulin Quantum Neural Network A quantum-inspired computational substrate.
The system converts structured numeric data into high-dimensional inference embeddings and phase-based decision outputs.
⚠️ The Core substrate is NOT provided in this repository.
Unlike classical ML systems:
No training loops
No gradient descent
No weights or tuning
You send structured data → the substrate returns:
inference probabilities
activation threshold (tau)
qualia embedding snapshot
intent vector
decision geometry
The engine behaves as a quantum-inspired inference oracle, not a train-and-predict model.
🧠 Functional Modes
EEG
Input: (channels × samples) matrix Output: state basins, coherence bands, intent vectors
Tabular
Input: row-wise numeric samples Output: class basin, phase threshold, decision quality
Finance
Input: OHLCV, indicators, rolling features Output: directional probability, phase confidence, action geometry
Image (Beta)
Input: flattened or tensor image Output: perceptual probabilities, embedding vectors
Any structured numeric array can be used as input.
🔑 Authentication
Customers receive an API key tied to a subscription tier.
Send via HTTP header:
x-api-key: YOUR_TQNN_API_KEY
Usage is tracked at runtime
Quotas are enforced
Overages are billed automatically
Core substrate logic is never exposed
📦 Client Installation
Official PyPI package (coming soon):
pip install tqnn-client
Until release, use the provided tqnn_client.py.
🚀 Quickstart Example — Tabular
from tqnn_client import TQNNClient import os
BASE_URL = os.getenv("TQNN_API_URL", "https://YOUR-TQNN-ENDPOINT") API_KEY = os.getenv("TQNN_API_KEY", "YOUR_KEY")
client = TQNNClient(api_key=API_KEY, base_url=BASE_URL)
data = [ [1.2, 0.4, 3.3, 0.1], [2.1, 1.1, 0.9, 0.5], [0.7, 0.3, 1.2, 2.1] ]
result = client.run_any( data=data, mode="TABULAR", label="demo_table" )
print(result)
📫 API Response Format
Example:
{ "mode": "TABULAR", "label": "demo_table", "probs": [0.18, 0.44, 0.38], "threshold": 0.613, "qualia": "...", "intent": "...", "usage": 41 }
Field meanings
probs — inference probability distribution
threshold — activation score (tau)
qualia — substrate embedding snapshot
intent — decision geometry
usage — runtime quota counter
📂 Repository Contents
Included:
API wrapper utilities
SDK client code
Request/response schemas
Integration examples
Public demos
Not included:
Core substrate
Inference algorithms
Architectural models
Internal runtimes
🛡️ Licensing
This project is dual-licensed.
✔️ MIT License — Open Layer
Applies to:
API wrapper
Integration libraries
SDK utilities
Example scripts
Public demos
You may freely:
Use
Modify
Integrate
Redistribute
See LICENSE.
🔒 Proprietary License — Core IP Locked
The following are closed-source and protected:
TQNN Core Engine
Tubulin substrate architecture
Quantum-inspired inference substrate
Qualia & intent embeddings
Internal runtimes and training pipelines
Access requires:
Paid subscription or
Enterprise licensing agreement
See TQNN-Core-License.md.
This repository contains:
Public API endpoints
Client utilities
Integration examples
This repository does not contain:
Substrate logic
Inference circuits
Decision models
Architectural mechanisms
Attempts to:
reverse-engineer
simulate substrate behavior
train competitor models using embeddings
reconstruct internal logic
constitute trade secret infringement.
💳 Billing Model
Tier Monthly Requests Intended Use
Tier 1 10,000 Builders / Research Tier 2 50,000 Startups / Teams Tier 3 200,000 Enterprise / Multi-modal
After quota exhaustion:
Requests continue
Per-unit billing applies
Core engine remains sealed
🗺 Roadmap
PyPI client package
CLI tooling
Multi-modal SDK modules
Android edge inference
Enterprise substrate clusters
GPU acceleration
📬 Contact
Enterprise licensing & integration: tqnnlabs@gmail.com
Final Reminder
This repository provides:
API surface
Client utilities
Usage examples
It does not provide:
The substrate
The inference models
The architecture
The Core remains sealed.
Built by TQNN Labs — A solo research effort. Contact: tqnnlabs@gmail.com
🔗 Live Endpoint (Coming Soon) https://api.tqnn.dev