Category: Tool update

CorVpoint STUDIO Private AI: AI under your control

Private AI brings local models, protected knowledge spaces, document analysis, translation, speech, CorVpoint Tools and governed online providers into one central workspace.

Published: September 22, 2026

Overview: CorVpoint STUDIO Private AI with a central launchpad, CorVpoint Tools, local AI, online AI and tool packagesPrivate AICentral launchpadCorVpoint ToolsLocal AIOnline AITool packages
In development

CorVpoint STUDIO Private AI is a locally oriented and modular AI workspace. It connects models, tools, data and user permissions in one central launchpad. The platform always shows whether a capability runs locally, in a governed hybrid mode or online.

Models and providers are managed as separate, controlled components. Private AI supports Ollama as well as OpenAI-compatible and vLLM-compatible runtimes. Guided checks distinguish installation, running service, available model, provider, deployment, policy and a real functional test before a capability is shown as ready.

PDF, DOCX, XLSX, TXT, Markdown, CSV and safe HTML files can be added to protected knowledge spaces. Search combines keywords with local features, respects access permissions and returns source references and trust labels. Roles, groups and permissions determine who can view and use information.

Daily work includes several intelligence modes, chat with multiple models, local conversation search, summarization, governed translation, speech-to-text, agent templates and specialized workspaces. Data protection rules can detect sensitive content and allow, report, mask or block it according to policy.

CorVpoint Tools and reviewed local packages can be previewed, checked for integrity, installed, started, stopped and removed. Plugins, external tools and online providers remain clearly separated. Installation or visibility never grants execution permission automatically.

Secure operation includes local authentication, session management, audit evidence, governed network profiles, encrypted backups, staged restore and transparent readiness views. Optional components such as local speech recognition use guided steps with progress, cancellation, recovery and automatic reconciliation of provider, model and policy.

The current development state has been tested extensively on the native development host. Actual readiness depends on installed runtimes, available models, permissions, policies and licenses. Private AI makes that state visible and remains a development candidate until the open signing and fresh-system gates are complete.