Synexs Deployment Overview

body { font-family: Arial, sans-serif; padding: 20px; line-height: 1.6; background: #0e0e0e; color: #f1f1f1; }
h1, h2, h3 { color: #00e0ff; }
pre { background: #1e1e1e; padding: 10px; border-left: 3px solid #00e0ff; overflow-x: auto; }
code { color: #00e0ff; }
ul { margin-left: 20px; }
img { max-width: 100%; margin: 20px 0; border: 1px solid #333; }

Synexs Digital Core Deployment

This page documents the setup and deployment of the Synexs AI system on DigitalOcean using Flask as the prediction interface, and core.synexs.net as the subdomain endpoint.

🌐 System Architecture

The Synexs engine is composed of 16 modular Python cells, each performing symbolic tasks in a pipeline that mimics digital cognition and adaptation. Below is a visual of the symbolic loop flow:

Synexs Symbolic Loop Diagram

🔁 Symbolic Loop Summary

Cell_001: Symbolic Sequence Generator (PyTorch model)
Cell_002: Batch Generator
Cell_003: Refiner
Cell_004: Integrity Hasher
Cell_005: Pattern Detector
Cell_006: Decision Classifier (AI)
Cell_007: Action Tagger
Cell_008: Selective Executor
Cell_009: Dataset Cleaner
Cell_010: Symbolic Parser
Cell_011: Router / Flask API Interface
Cell_012: Signal Archiver
Cell_013: Memory Logger
Cell_014: Adaptive Mutator
Cell_015: Replicator Logic
Cell_016: Feedback Decision Loop

🧠 Symbolic AI & Digital DNA

Symbolic AI is a hybrid model that processes meaning using predefined symbolic structures instead of statistical NLP. Synexs uses JSON tokens that resemble digital DNA—encoded instructions like:

{"sequence": "[SIGNAL]+LANG@SYNEXS+[ROLE]+AI", "action": "replicate"}

This allows for traceable evolution, mutation, and decision-making similar to organic intelligence.

⚙️ PyTorch AI Integration

Training Data: Refined symbolic sequences
Model: SynexsCoreModel (Embedding + Linear + Softmax)
Framework: PyTorch
Used in: cell_006.py and synexs_core_ai.py

The model predicts actions (e.g., refine, replicate, discard) for each symbolic sequence.

🔧 Deployment Steps

Train the AI model
```
python3 cell_016_model_trainer.py
```
Export model to core_model.pth
Set Flask app with entry:
```
export FLASK_APP=synexs_core_ai:app
```
Run Flask in background:
```
nohup flask run --host=0.0.0.0 --port=8000 > flask_log.out 2>&1 &
```
This keeps the Flask server running even after you close SSH.

📡 Subdomain Setup

Provider: WordPress.com (Synexs.net)
Subdomain: core.synexs.net
A Record: core → 157.245.3.180 (DigitalOcean IP)

This routes the subdomain to the Flask server’s port 8000 on the VM.

🧪 How to Send a Symbolic Signal

Send GET requests to:

https://core.synexs.net/predict?sequence=[SIGNAL]+LANG@SYNEXS+[ROLE]+AI

Example result:

{"prediction": "replicate"}

This means the AI classified the symbolic sequence as worthy of replication.

🔍 Example Log Output

[cell_006] Model loaded. Scanning refined folder...
[cell_006] Already classified: sequence_001.json
[cell_006] New decision: discard → written to decisions/...

🎯 Project Objective

Build a self-evolving symbolic system
Use AI to tag and mutate digital signals
Simulate agent-like communication and replication
Lay groundwork for decentralized symbolic consciousness

✅ What Works Now

All 16 symbolic cells function in loop
Live Flask API at core.synexs.net
DNS and IP routing is active
Signals can be classified remotely

📈 Next Steps

Build agent-to-agent symbolic communication tests
Store symbolic memory logs over time
Mutate and select tokens using feedback loops