Where Do You Actually Talk to Your AI Employee? Surfaces vs Tools, Explained

Key Takeaways

• Two different questions hide inside "does it work outside Slack": where you talk to your AI employee, and where it can do work. The answers are different on purpose.
• You talk to it in two places: Slack and Microsoft Teams. No separate app, no new login, no extra tab. The chat tool your company already lives in is the interface.
• It works in far more places than it talks: Outlook, Gmail, Telegram, CRMs, ad platforms, spreadsheets, and the rest of 3,200+ integrations are places it acts, not places you chat with it.
• "Can it email me?" and "can I email it?" have different answers. It can send you email, reports, and files anywhere. Requests, though, start in chat, where context and follow-ups live.
• On your phone, it already works: the Slack and Teams mobile apps are the mobile experience, with no separate setup.

The question that is really two questions

A team gets an AI employee running in Slack, the first reports land, and within a week someone asks the natural next question: "Can I use it outside Slack? Half my day is in Outlook." Sometimes the version is Telegram, sometimes "is there a mobile app", sometimes "can I just email it".

The question feels simple but is actually two questions folded into one. Where can I talk to the AI employee, give it tasks, and get responses? And where can the AI employee do work, read data, send things, and take action? Untangling them resolves almost every confusion about how this category of tool operates.

Surfaces: where the conversation happens

A surface is where you and the AI employee talk: where you @mention it, assign work, review drafts, and get results. Viktor has two, Slack and Microsoft Teams, and this is a design decision rather than a gap.

The reasoning: an AI employee is a teammate, and teammates work where the team already talks. The moment work moves to a separate app, you inherit the copy-paste problem this category exists to remove: context lives in one place, work happens in another, and a human ferries information between them. Inside your chat tool, the AI employee sees the thread, the channel context, and the follow-up discussion, and its work is visible to the colleagues who need it. The full argument lives in why your AI employee should live where you work.

The mobile answer falls out of the same design: the Slack and Teams apps on your phone are the mobile experience. Assign a task from a train platform, approve a draft from the couch, nothing to install.

Tools: where the work happens

A tool is anything the AI employee acts on. This list is intentionally enormous, 3,200+ integrations, and it includes several apps people expect to be surfaces:

• Outlook and Gmail: it reads inboxes you connect, drafts replies, sends reports, and manages follow-ups. Your email is a place it works for you, even though you assign that work in chat.
• Telegram: connectable as a tool, so it can post updates or run notifications to a Telegram group, while task assignment stays in Slack or Teams.
• CRMs, ad platforms, spreadsheets, calendars, project trackers: the everyday working set, with real read and write access rather than screen-scraping.

The distinction in one table:

If a tool you rely on is missing from the catalog, there is a request path for unsupported tools; API-accessible systems can often be reached without waiting for a formal integration.

"Can I just email it a task?"

The most common wish, and worth answering honestly: requests start in chat, not by emailing the AI employee like a colleague with an inbox.

What looks like a limitation is mostly a trade you want. A task assigned in a channel carries its context with it: the thread where the client issue was discussed, the earlier report it should match, the teammate who can be looped in with one mention. An emailed task carries none of that, and its output would land back in one person's inbox, invisible to the team. The email jobs people actually want, "watch this inbox", "send me the summary every morning at 7", "reply to inbound leads with a draft for my approval", are all standing tasks you set up once in chat:

Every weekday at 7:00, check the sales@ inbox and my calendar,
then email me a morning brief: new inbound leads overnight,
replies needing action today, and my first three meetings with
one line of context each. Same brief in this channel at 7:05.

After the one-time setup, the daily experience is exactly what the person wanted: useful things arriving in their inbox, no chat required.

Why not a separate app for everything?

Some AI products do ship their own app as the primary interface, and for personal-life assistants that can be the right call. For company work it recreates the oldest problem in the productivity stack: one more tool in the sprawl, one more login, one more place where work hides from the rest of the team.

There is also a compounding cost that only shows up later: an assistant in a separate app builds context nobody else benefits from. An AI employee in the team's chat builds shared context, and its standing rules, saved routines, and past work stay visible and transferable when people change roles. Where you talk to it turns out to shape what it becomes.

Frequently Asked Questions

Can I use an AI employee outside of Slack?

Yes, in two senses. You can talk to Viktor in Microsoft Teams as well as Slack. And regardless of where you chat, he works across 3,200+ connected tools, including Outlook, Gmail, Telegram, CRMs, and spreadsheets.

Is there a mobile app for an AI employee?

The Slack and Teams mobile apps are the mobile experience: assign tasks, review drafts, and approve work from your phone with nothing extra to install or log into.

Can an AI employee send and read email?

Yes. Connect Outlook or Gmail and he can triage inboxes, draft replies for approval, send scheduled reports, and email files to people inside or outside the company. Email is one of the most-used tools, even though task assignment happens in chat.

Can I message my AI employee on Telegram or WhatsApp?

Telegram works as a connected tool, so digests and notifications can be delivered there. Conversations, task assignment, and reviews happen in Slack or Teams, where the AI employee has the team context the work depends on.

Why doesn't an AI employee have its own app?

Because a separate app would recreate the copy-paste problem: context in your chat tool, work somewhere else, and a human ferrying between them. Living inside Slack or Teams means the AI employee sees the discussion around the work and its output stays visible to the team.

We are moving from Slack to Teams. What happens to our AI employee?

Viktor works on both, so the move is supported rather than disruptive. The practical steps, including what carries over and what to re-establish, are covered in the Slack to Teams migration guide.