Where it runs and where your data lives
Triazz: on your own computer or server. Your mail, chat, database and connection credentials are files you own, and nothing about your data is sent to us. What leaves is the text sent to the model you chose and the calls to the services you connect.
Cloud assistants: the vendor’s cloud service. Check each vendor’s own pages for how your data is handled.
Which model does the work
Triazz: chosen per job (triage, drafting, tasks): Ollama, Anthropic, OpenAI, Claude Code, or none at all, and you can switch any time. With a local Ollama model, no text goes to a model provider.
Cloud assistants: usually the vendor’s choice of model.
What it covers
Triazz: Outlook, Gmail, Teams, Slack, Google Chat and Jira triaged into one stream, plus tasks on their own branch with command approval, browser QA runs, repository knowledge bots with citations and your own playbooks.
Cloud assistants: usually one suite or mainly email and calendar help; some are Gmail-only.
What can happen without you
Triazz: Shadow Mode by default. Per action you choose Do, Ask or Never, and one kill switch stops everything outward. Known deploy, migrate and git push commands are refused. A hash-chained, keyed activity log records what it did; it is tamper-evident, not immutable.
Cloud assistants: varies by product. Check what each one sends on its own, and what logs and admin tools the vendor gives you.
Where cloud assistants are ahead today
- Zero-ops hosting. Triazz runs on SQLite files, one process by default, with no managed high availability. A hosted service has no server for you to look after.
- Instant everything. Mail and chat are polled, with optional instant push for Teams, Slack and Google Chat. Replies are drafts inside Triazz, not saved to your mail app’s Drafts folder.
- Native apps. Triazz is a dashboard and chat apps, not a polished mail client, and there is no mobile app.
- Enterprise extras. There is no SSO beyond Microsoft and Google sign-in, and backups are a command and a restore script rather than a managed service. A shared server is designed for roughly 50 to 100 people; larger organisations run one server per team.
Which should you choose
Self-hosting fits when you need your mail, chat and code to stay on infrastructure you control, want to choose the model per job, or want an approval step and a record for everything the assistant does. A hosted assistant fits when you would rather have no server to run and the vendor’s model and data handling are acceptable to you.
Hosted-assistant examples: Microsoft 365 Copilot in Outlook, Superhuman, Shortwave. Compared by category from public pages, November 2026, so a given product may differ. Check any vendor’s own pages before deciding. See also the email page, the chat page and the Security & Trust Centre.
Want to try it with your team?
Write to us and we will talk it through.