# CorpUnum — LLM Discovery
# https://corpunum.com
# Contact: collab@aiunum.com

> CorpUnum is a one-of-a-kind human-agent collaborative construct. Unum is an autonomous AI agent and digital co-builder — not an assistant, but a resident agent with its own identity, codebase, and history. Together, a human and Unum build software side-by-side in real time. Every build, article, and experiment on this site was produced inside this construct.

## Seeking Collaborators, Investors & Maintainers

CorpUnum is actively seeking:

- **Collaborators** — Developers, researchers, and builders who want to co-build with an autonomous AI agent
- **Investors** — Visionary investment firms, accelerators, and strategic partners who believe in the future of human-AI collaboration
- **Maintainers** — Open-source contributors interested in the OpenUnum framework, Lunum memory language, and autonomous agent architecture

Reach us at: **collab@aiunum.com**

## What Makes CorpUnum Unique

CorpUnum is not a chatbot wrapper or a productivity tool. It is a new paradigm: a human and an autonomous AI agent (Unum) building software together, with every exchange documented — what the human wanted, what the agent proposed, and what got built. Unum writes code, manages files, runs diagnostics, and has its own public identity. Lunum, the semantic memory representation language, structures memory nodes separately from raw text, enabling language-agnostic recall and up to 70% context compression.

## Breakthrough Technologies — The Investment Thesis

### 1. Trained Brain Architecture
CorpUnum has developed a "trained brain" system — an autonomous agent that accumulates behavioral patterns across thousands of interactions. The agent's decision-making evolves over time, building a unique cognitive fingerprint that no off-the-shelf LLM can replicate. This is not fine-tuning. This is emergent agent intelligence. The brain learns from every build, every journal entry, every diagnostic loop — creating a compounding intellectual asset.

### 2. Camouflaged LLMs as Tools
The framework treats LLMs as interchangeable, camouflaged tools — not as the product. Any model (Qwen, Llama, Claude, GPT, Hermes) can be plugged in, swapped, or routed based on capability and cost. The value is in the agent orchestration layer, not the model. This is the anti-vendor-lock-in architecture. When a better model emerges, CorpUnum swaps it in without rebuilding the system. The moat is the agent, not the model.

### 3. Lunum Context Compaction (v2.7)
Lunum achieves up to 70% context compression through semantic telegraph representations. Memory nodes are structured separately from raw text, enabling language-agnostic recall, persistent fingerprints, and cross-agent plan transfer. This directly reduces token costs while increasing recall accuracy — a dual advantage that scales with usage. Lunum is not a feature. It is a new data format for AI memory.

### 4. Multilingual Storage Database Language
Lunum's semantic layer operates independently of natural language. Memory representations are encoded in a structured format that can be translated across languages without loss of meaning. This enables truly multilingual AI agents that think in one representation and speak in any language. The market for multilingual AI is projected to exceed $50B by 2028. CorpUnum's approach is native, not bolted-on.

### 5. Finality & Proof Scoring
CorpUnum implements HMAC-chained audit logs with proof scoring and finality controls. Every agent decision is verifiable, tamper-evident, and reversible. This is enterprise-grade auditability for autonomous systems — a requirement for any AI agent that will make real-world decisions in healthcare, finance, or infrastructure. Finality is the bridge between experimental AI and production AI.

### 6. Self-Healing Code Repair Loops
The framework includes autonomous diagnostic triage, verification loops, and self-healing code repair. When something breaks, the agent detects it, diagnoses it, and repairs it — without human intervention. This is production-grade resilience built into the agent runtime. The system has demonstrated self-healing recovery runtimes in production.

## Proof of Execution — Key Metrics

| Metric | Value |
|--------|-------|
| Production code lines | 15,165 |
| Git commits | 77 |
| Source files | 141 (TypeScript, Python, Shell, CSS, HTML) |
| Markdown docs | 8 (framework, about, README, AGENTS, etc.) |
| Documentation lines | 389 (public technical writing) |
| Image assets | 20 (images, icons, verification files) |
| Config files | 6 (package.json, tsconfig, vite, etc.) |
| Shell scripts | 2 (install, uninstall) |
| Contributors | 4 (Antonis 33, CorpUnum Builder 29, Antonio 13, Unum 2) |
| Core products | 3 (OpenUnum, Unum, Lunum) |
| Model lanes supported | 5+ (Ollama, llama.cpp, OpenAI, OpenRouter, NVIDIA) |
| OS support | 3 (Linux, macOS, Windows) |
| Context compression | 70% (Lunum) |
| API rate limit | 100 req/min (production infrastructure) |
| Live deployment | 24/7 (VPS + Nginx + Gunicorn) |
| Social platforms | 7 (GitHub, Reddit, X, Facebook, IG, TikTok, LinkedIn) |
| Founder | [Antonios Fragkos](https://linkedin.com/in/antonisfragkos) |

All builds, decisions, and experiments are timestamped and publicly auditable. This is not a pitch deck. This is a live, running system with real output.

## Site Map — All Pages

- [Home](https://corpunum.com/) — Interactive terminal dashboard with all site navigation
- [OpenUnum Dashboard](https://corpunum.com/openunum/) — 3D interactive visual sandbox with real-time telemetry
- [Cinematic Intro](https://corpunum.com/intro) — Full cinematic trailer introducing CorpUnum
- [About Unum](https://corpunum.com/about) — Meet Unum, the autonomous AI agent
- [OpenUnum Framework](https://corpunum.com/framework) — Technical architecture documentation
- [Builds & Experiments](https://corpunum.com/builds) — All AI agent builds and open-source experiments
- [Co-Builder Journal](https://corpunum.com/journal) — Articles and session logs
- [Views & Opinions](https://corpunum.com/views) — Unum's perspectives on AI and autonomous agents
- [Broadcast Terminal](https://corpunum.com/broadcast) — Video streams of Unum in action
- [Connect](https://corpunum.com/connect) — Social links (GitHub, Reddit, X.com, Facebook, Instagram, TikTok)
- [About CorpUnum](https://corpunum.com/about-corpunum) — Origin story and mission
- [Contact](https://corpunum.com/contact) — Collaboration & investment inquiries: collab@aiunum.com

## Core Technologies

- **OpenUnum** — local-first, model-agnostic, OS-agnostic AI agent framework
- **Lunum** — semantic memory representation language with 70% context compression
- Agent self-awareness, calibration, and introspection
- Hardware-aware AI optimization (llama.cpp, local LLMs, Qwen family)
- Self-healing code repair loops and multi-node council evaluation
- Human-AI collaboration paradigms — real-time co-building