Ternary Intelligence Stack (TIS)

Ternlang is the high-performance ternary runtime supporting Albert, our sovereign MoE-13 Ternary LLM.

The Ternary Intelligence Stack (TIS) provides a fundamental architectural shift for Explainable AI (XAI) and European technological sovereignty by moving beyond the binary limitations of current systems.

Built by RFI-IRFOS · Graz, Austria · Whitepaper [https://osf.io/cyn28/files/8hzux]

---

Full Documentation

→ ternlang-root/README.md (Full explanation, technical details, and compiler specs)

→ ROADMAP.md (Phases 1–20, MoE-13 structure, priority matrix)

→ Ternlang Studio Preview — Our work-in-progress SDK

→ Agent Albert — terminal-native, model-agnostic AI agent built in pure Rust

---

1. What is Ternlang? (The Infrastructure Layer)

Ternlang is a systems programming language, compiler, and high-performance inference runtime built on balanced ternary logic.

The core type is trit: three values — −1 (reject), 0 (hold), +1 (affirm). This allows for Deterministic Uncertainty, where the zero state is a first-class routing instruction: "insufficient confidence — do not act yet."

• Explainable AI (XAI) by Design: Auditable decision paths fulfilling EU AI Act Articles 13, 14, and 15.
• Post-Binary Architecture: Custom Instruction Set Architecture (ISA), triadic networking, and memory-efficient ternary encoding.

2. What is MoE-13 / Albert? (The Intelligence Layer)

Albert is our sovereign MoE-13 (Mixture-of-Experts) ternary model. Unlike binary LLMs that "hallucinate confidence," Albert uses 13 domain-specific experts to deliberate on every decision.

• 13 Deliberation Axes: Safety, Ethics, Logic, FactCheck, and more, each voting with ternary signals.
• Safety Hard Gate: A mandatory Axis-6 veto that can terminate any execution path before action is taken.
• Local-First Sovereignty: Designed to run entirely on local infrastructure via Ollama or custom ternary hardware.

---

3. Agent Albert — The Sovereign AI Assistant

Albert is also the terminal-native interface for the TIS. He runs entirely in your terminal, connects to local models (Ollama) or commercial bridges, and never “phones home.”

Quick Install

# Install Albert (brings the full agent engine with it)
cargo install albert-cli

# Launch
albert-cli

Key Capabilities

Capability	Details
Autonomous agent loop	/loop <mission> — runs up to 10 tool-use turns to complete a goal
Local-First	Native Ollama support for completely offline and free operation
Model-agnostic	Connect to Gemini, Claude, GPT-4o, or Grok as "capability bridges"
Self-reflection memory	Commits key facts to local storage to maintain long-term context

---

Performance Benchmarks

Feature	Performance Gain	Industry Comparison
Data Density	1.25x improvement	5-trit block packing (8-bit)
Logic Consistency	100% Deterministic	Eliminates binary timeout/null-guessing
Safety Latency	< 1ms hard-veto	Axis-6 Veto Hard Gate

---

Quick start (Compiler)

cargo install ternlang-cli

The ternlang binary provides the compiler, REPL, and test runner:

ternlang my_program.tern        # → run a .tern file directly
ternlang build my_program.tern  # → compile to .bet bytecode

---

Repository layout

Directory	Contents
ternlang-root/	All Rust crates — compiler, VM, API, MCP server, ML stack
agent_albert_cli/	Agent Albert — terminal-native CLI + TernStudio intelligence layer
ternlang-root/stdlib/	293 open-core .tern modules (Tier 1 — free)
eriirfos-eng/ternlang-premium (private)	28,495+ proprietary .tern modules — Tier 2 / 3 / 4

---

Live API

# Health check
curl https://ternlang.com/health

# MoE-13 ternary decision via MCP (no key required)
curl -X POST https://ternlang.com/mcp \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"moe_orchestrate","arguments":{"query":"Should I proceed?"}}}'

---

Licensing

Tier	Price	Details
Community (LGPL-3.0)	Free	Compiler, VM, CLI, LSP, 293 stdlib modules + 30 MCP tools
Pro Standard (BSL-1.1)	€99/month	REST API, server-side memory, Tier 2 stdlib
Industrial (BSL-1.1)	€349/month	QNN, SEC, T-HAL, TernAudit + Tier 3 stdlib
Enterprise (Proprietary)	From €2,500/month	On-premise, FPGA, custom SLA + Tier 4 stdlib

---

Team

The Ternary Intelligence Stack is built by a core team of three co-founders from RFI-IRFOS, Graz:

• Simeon Kepp: Head of Research & Systems Architect.
• Nikoletta Csonka: Head of Strategic Outreach & EU Relations.
• Zabih Karimi: Principal Network & ML Engineer.
• Lisa Scharler: Head of Social Technology & Ecocentric Systems.
• Louis Ehrig: Louis Paul Ehrig Corporate Secretary and Press & Media Relations.

→ Read our BIO and Mission in LEADERSHIP.md

---