betaUpdated February 16, 2026

Digital Employee Team

A team of five AI employees with real email addresses. Send them work — research, writing, decks, spreadsheets, project plans — and they'll deliver completed tasks with document attachments straight to your inbox. Track every workflow in real time from the dashboard.

Agentic AIMulti-AgentOpenAIAnthropicEmailVercel AI SDK
Digital Employee Team illustration

What it solves

Most executives have heard the pitch for "AI agents" but haven't experienced what a coordinated AI team actually looks like in practice. Slides about multi-agent orchestration don't land the same way as sending an email to casey@brianwhite.io and getting a polished PowerPoint deck back ten minutes later.

This tool makes agentic AI tangible by mapping it to a concept leaders already understand: an org chart with people who have names, titles, and email addresses. No new apps to learn. Just email — and a real-time dashboard to watch the work happen.

The team

  • Morgan (Chief of Staff) — Routes incoming requests to the right specialist, reviews every deliverable before it reaches you, and can handle quick questions directly. The manager of the operation.
  • Alex (Research Analyst) — Deep web research, competitive analysis, market briefs, and data-backed summaries.
  • Jordan (Content Strategist) — Long-form writing, editing, content strategy, and Word document generation.
  • Casey (Data & Deck Builder) — PowerPoint presentations, Excel spreadsheets, Word documents, and data analysis.
  • Riley (Operations Coordinator) — Project plans, process design, operational docs, spreadsheets, and calendar invite (.ics) generation.

How it works

Behind the scenes, this is a multi-agent orchestration system powered by OpenAI and Anthropic models:

  1. Inbound email arrives at any team member's address (e.g. casey@brianwhite.io)
  2. Morgan routes the request to the right specialist — or handles it directly if it's a quick question
  3. The assigned agent works the task using web search, document generation (PowerPoint, Excel, Word), calendar invite creation, and persistent memory of your preferences
  4. Morgan reviews every deliverable and can send work back for revision before anything reaches you
  5. You get a reply with the finished work and any file attachments in your inbox

Each agent maintains persistent memory of your preferences, past interactions, and feedback — so the team gets better at working with you over time.

Workflow dashboard

Every task you send creates a workflow you can track in real time by viewing your Dashboard. The dashboard shows:

  • Live status tracking — Watch your request move through four stages: Queued, Team Working, Manager Reviewing, and Delivered
  • Agent activity timeline — See exactly which agent is working, when tasks are delegated, when work is submitted for review, and when revisions are requested
  • Deliverable downloads — Download generated files (PowerPoint, Excel, Word) directly from the dashboard
  • Auto-updating — The dashboard polls for changes every few seconds, so you can watch the team work without refreshing

Sign in to view workflows tied to your email address.

Ideal for

  • Executives and business leaders who want to experience agentic AI without learning a new tool
  • Teams evaluating AI who need a concrete, working demo of multi-agent coordination
  • Anyone who thinks in email and wants AI that meets them where they already work
  • Quick operational tasks — research briefs, first-draft decks, project plans, calendar invites, written content — that don't justify opening another app

Try it out

Meet the team

Want to adapt this pattern for your workflow? Share your context — feedback helps shape the next iteration.

Related Experiments