Skip to content

designlogic-robert/POCs

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

21 Commits
 
 
 
 
 
 

Repository files navigation

USS Proof-of-Concepts (POCs)

This repository contains small, focused Proof-of-Concept implementations demonstrating key ideas from the Universal Semantic Systems (USS) architecture.
Each POC is self-contained, easy to run, and illustrates how USS enables structured reasoning, deterministic planning, and semantic control on top of foundation models.


📁 Included POCs

1. caller-uss-poc/

A minimal, executable demonstration of CALLER, the USS request-orchestration layer.

This POC shows:

  • How a request enters the system through a structured message schema
  • How CALLER interprets intent and extracts required tasks
  • How USR (Universal Semantic Runtime) creates a deterministic plan
  • How small “engines” produce structured, auditable outputs
  • How CALLER assembles everything into a clean, final answer

Why this matters:
It demonstrates the USS pattern: structured reasoning → controlled planning → reproducible, non-hallucinatory outputs.

Files:

  • caller_poc.py — fully executable Python POC
  • caller_overview.md — concept and architecture explanation
  • caller_walkthrough.md — step-by-step example flow
  • README.md — usage instructions

Run with:

python caller_poc.py

2. synce-poc/

This POC demonstrates a minimal, self-contained simulation of the core USS execution pattern using SynCE-style components.
It is not a full runtime — it’s a compact, readable demo showing how semantic tokens, SCP routing, deterministic planning, and engine execution interact.

Perfect for founders, recruiters, or collaborators who want to see how USS behaves in code.


What This POC Shows

✅ Semantic Tokens

How raw input becomes structured semantic units (tokens.py).

✅ SCP (Semantic Control Protocol) Routing

How SCP interprets tokens and selects appropriate actions (scp_layer.py).

✅ Deterministic Planning (ORCH-C Mini)

A tiny version of ORCH-C that turns intent into a linear, explicit execution plan (orchc.py).

✅ Engine Execution

main.py runs the full cycle:

  1. Ingest raw input
  2. Tokenize
  3. SCP routing
  4. ORCH-C planning
  5. Engine execution
  6. Output assembly

✅ Example Output

example_output.txt shows what a complete run looks like.

This demonstrates the signature properties of USS:

  • No chain-of-thought improvisation
  • Explicit intermediate states
  • Deterministic reasoning path
  • Clear separation of layers

File Overview

synce-poc/
│
├── docs/ # (Optional) project notes / architecture diagrams
│
├── main.py # Entry point – runs the full POC
├── tokens.py # Semantic token definitions
├── scp_layer.py # Semantic Control Protocol logic
├── orchc.py # Deterministic planner (mini ORCH-C)
├── example_output.txt # Sample output from a full run
├── requirements.txt # Dependencies (very lightweight)
│
├── LICENSE # Apache 2.0 (open + safe to use)
└── README.md # This file

Running the POC

1. Install dependencies

pip install -r requirements.txt
  1. Run the demo
python main.py

You’ll see a full USS-style reasoning flow printed to the console.

Why This Matters

  • This POC is intentionally tiny, but it demonstrates:
  • How USS structures meaning before reasoning
  • How semantic tokens drive deterministic planning
  • How SCP + ORCH-C remove hallucination drift
  • How the system remains fully inspectable, step-by-step

Anyone reviewing this POC gets a clear sense of what makes USS different from “just another LLM wrapper.”

🌐 Purpose of This Repository

These POCs help:

  • founders
  • technical leads
  • recruiters
  • collaborators
  • platform architects

…quickly understand the practical value of USS through compact, observable examples.

They are not production code. They are clarity tools showcasing how USS behaves in small, testable environments.

🧱 What USS Demonstrates Across These POCs

  • Deterministic orchestration above any LLM
  • Layered semantic architecture (Tokens → SCP → USR → Engines)
  • Explicit, machine-auditable reasoning
  • No improvisational chain-of-thought
  • Stable, reproducible outputs
  • A universal pattern for multi-domain reasoning systems

📬 Contact

Robert Hansen Chief Semantic Architect — USS

🔑 License

All POCs in this repository are released under Apache 2.0 unless otherwise noted.

Releases

No releases published

Packages

No packages published

Languages