README / README.md
dcata004's picture
Update README.md
9f1571d verified
|
raw
history blame
3.89 kB
---
title: README
emoji: 🛡️
colorFrom: blue
colorTo: red
sdk: gradio
pinned: false
sdk_version: 6.2.0
---
<div align="center">
# 🛡️ Cata Risk Lab
### Sovereign AI Assurance & Technical Gatekeeping
**Miami | Serving USA, Swiss, UK & Australian Markets**
[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
[![DOI](https://zenodo.org/badge/DOI/10.5281/zenodo.18174088.svg)](https://doi.org/10.5281/zenodo.18174088)
[![Compliance: nFADP](https://img.shields.io/badge/Compliance-nFADP%20Ready-red)](https://www.fedlex.admin.ch/eli/cc/2022/491/en)
[![Compliance: EU AI Act](https://img.shields.io/badge/Compliance-EU%20AI%20Act-blue)](https://artificialintelligenceact.eu/)
<br/>
> *"In an era of generative hallucination, a guess is not a glitch. It is a lawsuit."*
</div>
---
## 🌍 Open Source Mission
Cata Risk Lab builds **public infrastructure** for AI sovereignty. We believe compliance tools should be open goods—not luxury services sold behind closed doors.
Our mission: Empower any developer to run enterprise-grade sovereignty audits in minutes, democratizing capabilities previously reserved for Fortune 500 companies.
**Built with care. Powered by people. Open by design.**
We do not build "creative" chat bots. We build **Sovereign Auditors** that enforce:
* **Data Sovereignty:** Ensuring data never crosses borders illegally (nFADP/GDPR).
* **Citation Accuracy:** If the AI cannot cite the source, the AI must fail.
* **Liability Containment:** Shifting risk away from the enterprise.
---
## 🏛️ Deployed Sovereign Tools (Live)
### 1. [🇨🇭 Swiss Risk Calculator](https://huggingface.co/spaces/Cata-Risk-Lab/Swiss-Risk-Calculator)
* **Role:** Regulatory Risk Assessment
* **Function:** Maps AI deployments against the **Swiss nFADP** and **EU AI Act**. Calculates risk tiers based on data sensitivity and user volume.
### 2. [🛡️ Veritas: The Hallucination Auditor](https://huggingface.co/spaces/Cata-Risk-Lab/Veritas-Auditor)
* **Role:** RAG Auditing Engine
* **Function:** Ingests financial PDFs and forces strictly cited responses. Returns `FAIL` verdicts for hallucinations using the Ragas framework.
### 3. [🇦🇺 Wattle Guard: Australian Sovereignty](https://huggingface.co/spaces/Cata-Risk-Lab/Wattle-Guard)
* **Role:** APP 8 Compliance Validator
* **Function:** Automates cross-border disclosure audits for Australian entities. Detects server jurisdictions to prevent SOCI Act violations.
### 4. [📜 Liability Scanner](https://huggingface.co/spaces/Cata-Risk-Lab/Liability-Scanner)
* **Role:** Contract Forensics
* **Function:** Scans vendor "Terms of Service" URLs for high-risk clauses (e.g., "Right to Train," "Perpetual License") that violate data privacy.
---
## Industrial Tech Stack
We build strictly on enterprise-grade, auditable infrastructure.
| Layer | Technology | Purpose |
| :--- | :--- | :--- |
| **Orchestration** | `LangChain` / `LlamaIndex` | Logic flow and chain-of-thought governance. |
| **Inference** | `Azure OpenAI (GPT-4o)` | Private, firewalled reasoning engine. |
| **Memory** | `ChromaDB` / `Azure AI Search` | Vector storage with strict access controls. |
| **Deployment** | `Hugging Face Spaces` | Public audit infrastructure. |
| **Validation** | `CERN Methodology` | Reproducible testing protocols for AI outputs. |
---
## Contact & Governance
**Cata Risk Lab** operates as a technical due diligence partner. We provide fixed-price audit packages for Shadow AI Discovery and Vendor Due Diligence.
* **Principal:** Anthony D. Cata
* **Locations:** Miami (HQ) | Zurich (Partner Network)
* **GitHub:** [github.com/dcata004](https://github.com/dcata004)
* **Verification:** All software artifacts are DOI Minted via Zenodo for permanent audit trails.
<div align="center">
<br/>
<i>"Sovereignty is the new Compliance."</i>
</div>