The Keys Systems announces a working private AI pilot for questions and answers over selected files. The initial implementation has been tested on an NVIDIA RTX 4070 Laptop GPU locally and on a Tesla T4 in a separate private cloud environment.
A focused first version
The pilot brings together a private workspace, the Qwen3:4b model, Ollama, and container tooling. Its initial purpose is specific: help its owner work with selected knowledge while retaining control over what is added to the workspace.
Local and cloud inference were exercised during deployment verification. The cloud verification used a synthetic test document. The cloud service is accessed privately and operated on demand; this announcement does not offer an always-on public AI endpoint.
Deliberate growth
The pilot is the first practical expression of the TKS Tri-Core approach: think clearly about the task, challenge the answer, and connect the tools needed to carry the work forward.
Future development will be guided by concrete use cases, evaluation, appropriate access controls, and explicit cost decisions. Direct OpenAI, Gemini, and Claude model API connections remain separate integration work. The current pilot does not provide quantum computation or replace the Windows operating system.
Technology context
NVIDIA hardware is part of the tested deployment. NVIDIA AI Workbench provides the local development environment, with its documented support for local and remote GPU work. Learn more in NVIDIA’s AI Workbench documentation.
This is a TKS development announcement. It does not announce a formal NVIDIA partnership, sponsorship, certification, or endorsement. Product names belong to their respective owners.
Explore the pilot
Read about the TKS platform or contact TKS to discuss a focused private AI use case.
About The Keys Systems
The Keys Systems is an independent AI and technology initiative led by Lado Rigvava. Its work focuses on private AI, connected tools, and human direction.