AI products should reach the desktop when the work lives there.

Not every serious workflow belongs in a browser tab.

Many high-value workflows involve local files, internal web apps, terminals, spreadsheets, documents, long-running desktop tools, and workstation-specific context. Desktop AI can give users a safer and more capable environment for work that crosses applications.

Dan helps teams design desktop AI products that connect local context, cloud APIs, model routing, tool permissions, secure storage, and human approval flows without turning the user's workstation into an unmanaged risk surface.

Desktop AI use cases

• File-aware assistants for documents, spreadsheets, PDFs, logs, code, and operational exports.
• Internal analyst workbenches that combine local context with backend APIs and AI services.
• Secure desktop tools for regulated users who need explicit approval before AI-generated changes.
• Developer productivity tools that understand repos, terminals, build output, and browser state.
• Desktop-to-cloud apps that use local UI while keeping central audit, authentication, and model routing.

Architecture questions

• What stays local, what goes to the cloud, and what should never be sent to a model?
• How does the desktop app authenticate to enterprise services?
• How are file access, tool access, and user approvals logged?
• Which workflows need offline behavior or local inference?
• How should the UI expose AI suggestions without hiding risk?