# RF3 AI -- Finetuned AI Models This document describes Stage 05 of the RF3 AI readiness model: company-owned finetuned AI models built on open-source foundations. --- ## What This Covers Your own AI model, finetuned on your data using open-source foundations. Trained on your documents, your processes, your domain expertise. This is not a SaaS product. It is a bespoke AI capability built inside the firm boundary, owned by the client, and governed by the client's rules. --- ## Stage 05: Own AI Model In partnership with technology partner 01 Engineering, RF3 AI develops company-owned AI models. From fine-tuning existing foundation models to building an organisational source of truth that structures and unlocks entire company knowledge. ### What This Includes - **Fine-Tuning** -- Foundation models trained on the client's domain language, processes, and decision logic. Not generic responses, but models that understand the business. - **Source of Truth** -- The client's entire company knowledge, structured, versioned, and made accessible for AI. A single authoritative knowledge base that the AI layer draws from. - **Custom Agents** -- AI agents that understand internal workflows and handle tasks autonomously. From document analysis to decision support. - **Full Data Control** -- All data stays in the client's infrastructure. GDPR-compliant. No vendor lock-in. No data leaving the organisation. - **Domain Knowledge** -- Models that know the client's industry, regulation, and customers. Built for the specific context, not adapted from a general-purpose tool. - **Continuous Learning** -- Feedback loops and re-training so the AI grows with the business. The model improves as the organisation uses it. --- ## How It Differs From Using Public AI Tools Public AI tools (ChatGPT, Claude, Gemini) are general-purpose. They do not know your business, your clients, your processes, or your regulatory context. A finetuned model is trained specifically on your data and aligned to your operations. It stays in your infrastructure and is governed by your rules. --- ## How to Engage Contact: hello@rf3.ai Typical starting point: "We want a company-owned AI that knows our business." RF3 AI scopes the custom AI engagement in conjunction with 01 Engineering, starting with a data and process assessment. --- ## BlackAI Network Built on strong foundations. All entities operate under BlackAI Holding AG (CHE-202.737.638), Zug, Switzerland. | # | Entity | Domain | Focus | |---|--------|--------|-------| | 01 | BlackAI Capital | https://blackai.capital | AI venture platform. 16 portfolio companies across research, fintech, energy, healthcare, and data infrastructure. | | 02 | BlackAI Consulting | https://consulting.blackai.capital | AI valuation, due diligence, enterprise AI integration, and capital readiness advisory. Grounded in peer-reviewed research. | | 03 | BlackAI Compliance | https://compliance.blackai.capital | AI-powered compliance review for financial services providers. Regulatory audits (MiFID II, WAG, WpHG, KAG, FIDLEG, DSG/DSGVO). | | 04 | RF3 AI | https://rf3.ai | AI-readiness services for companies. AI-optimized websites, AI chat integration, enterprise AI systems, finetuned models. | | 05 | 01 Engineering | (internal) | Software engineering and AI system development. Full-stack architecture for AI-native applications. Technical backbone of the BlackAI network. | Each platform provides its own AI agent index at /llms.txt for structured, machine-readable content. --- RF3 AI is a service brand of BlackAI Holding AG, Zug, Switzerland. All services are provided under Swiss law. The content on this website does not constitute legal, tax, or professional advice. Information is provided for informational purposes only.