Donna

Human Interface

Check your email for your magic link

Donna

Checking...

This is a live production system. The protocols below are actually running right now — on schedules, reading real data, and taking real actions. The green dot above shows Donna's current status and how many primitives (operations) are available. For more context on what all this means, see the explainer at the bottom of this page.

Protocols

Recent Executions

Scheduled Automations

These run automatically on a schedule — no human needed.


What is a Protocol?

A protocol is a sequence of steps that Donna executes automatically. Each step calls a primitive — a small, focused operation like "read from a spreadsheet" or "generate text with AI." By chaining primitives together, protocols can do complex work reliably.

What can Donna do?

  • Call any major LLM — Claude, GPT-4, Grok, and more
  • Read and write data — Google Sheets, Notion databases
  • Fetch from APIs — HubSpot, custom endpoints, webhooks
  • Send notifications — deliver messages to your inbox
  • Run on schedule — hourly, daily, or custom timing

Why protocols instead of just asking AI?

AI is great for judgment and creativity, but unreliable for precise data work. Donna uses AI only where it adds value (writing, deciding, classifying) and uses deterministic code for everything else (merging data, formatting, calculations). This means your automations are consistent and debuggable.

✓

All caught up!

No pending messages from Donna

No message history yet

Hey!

Donna is an AI assistant that runs protocols on your behalf. Try one below to see how it works.

Jake's Automations

These are real protocols running in production right now. Each one is a multi-step workflow that Donna executes automatically — the same system you just used on the "Try It" tab.

Loading dashboard...

Each protocol can call any major LLM, query databases, send notifications, and more — all orchestrated by Donna's planner.