What is ShogunAI?
ShogunAI is a private AI memory assistant built for macOS. It captures the work context already living on your Mac — email, meetings, documents, and calendar — and turns it into a local memory you can query in plain language. Every answer arrives with its source and timestamp attached, so you can verify rather than trust.
Key capabilities
- Memory layer — Work across your apps is captured on-device and recalled with evidence, available offline.
- Contextual recall — One question spans mail, chat, docs, and calendar, with sources shown for each hit.
- Execution layer — Drafts, updates, and filings run inside the tools you already use, governed by three autonomy levels: reversible work runs on its own, drafts wait for you, and anything leaving your Mac stops for approval.
- Meetings — Transcripts become minutes with decisions and commitments; audio is never written to disk.
- Inline drafting — Press Option and a continuation is written at your cursor, locally, unsent.
- Live translation — Two-way translation between spoken languages, keeping only text afterwards.
- Daily brief — What moved overnight and what is still open, assembled before you ask.
Privacy by architecture
Memory and the text-search index live on your Mac. Visual screen memory is off by default with a retention period you choose. Only the specific slice a request needs ever leaves the device, and each transfer is written to a traceability log. You can run it on a supported assistant subscription or bring your own API key.
Who it is for
Knowledge workers, founders, product and engineering teams, and consultants who lose time reconstructing decisions and shuttling answers between tools. ShogunAI is macOS-only, built on native capabilities like the accessibility layer and system keychain, and runs on Apple Silicon. Plans open with a 7-day full trial.