Get Started

AI Assistant

Temway’s in-editor AI assistant is a chat drawer inside the email and layout editors. Describe what you want in plain language and it builds and refines your email on the canvas — adding blocks, rewriting copy, restyling elements, and applying your branding — while you watch each change land.

It runs on your own AI key (Anthropic, OpenAI, or any OpenAI-compatible provider), so the model and cost stay in your control. There’s no Temway-side metering, and it’s available on every plan.

For the why-it-matters angle and how it compares to connecting an external assistant, see The In-Editor AI Assistant.

How the assistant works

  • Runs on your key. You connect an Anthropic, OpenAI, or OpenAI-compatible key. The model is yours and usage bills to your provider account — not to Temway.
  • Your key is encrypted at rest. Once saved, the key is encrypted with AES-256-GCM and only decrypted on Temway’s server for the single moment a turn needs it. Your browser only ever holds a short, masked preview (the last few characters) — never the plaintext.
  • Keys are shared across the workspace. A key added by one editor is available to every member. You can keep several keys on hand and switch between them — and between models — from the drawer.
  • The assistant shares the canvas safely. It takes the edit lock on the email it’s working on, so its changes and yours never collide. You can stop a response mid-stream, take the lock back, and finish by hand whenever you want.

Adding your AI key

  1. Open Settings → AI Assistant.
  2. Pick a provider: Anthropic, OpenAI, or OpenAI-compatible (for example, a local Ollama install).
  3. Paste your key and save. Temway validates the key against the provider before saving — if it’s invalid or typoed, you’ll know immediately.
  4. The key is encrypted and stored. The workspace can now use it.

You can add more than one key (for example, an Anthropic key and an OpenAI key) and switch between them in the assistant drawer.

Using the assistant

  1. Open the assistant drawer inside the email or layout editor.
  2. Pick the key and model you want to use from the drawer’s selectors.
  3. Describe what you want in plain language — for example, “Build a welcome email from my branding: a hero headline, a short intro, one button linking to signup, and our social block at the bottom.”
  4. Watch each change land on the canvas. The assistant works one step at a time, and each step is listed in the chat.
  5. Refine in small steps — one change per message (“shorten the intro to two sentences”, “give the button a larger shadow”). Small steps are easy to review and undo.
  6. To interrupt, hit stop. The edit lock returns to you so you can finish by hand.
  7. When it’s ready, preview, then export the HTML or push to your ESP.

Providers

ProviderWhat to enter
AnthropicAn Anthropic API key from your console account.
OpenAIAn OpenAI API key from your OpenAI account.
OpenAI-compatibleAny provider that speaks the OpenAI API — including a local Ollama install. Enter the key (if any) and the base URL the assistant should call.

Tips

  • Set up branding first. “Use my branding” only works if your colors, fonts, and logo are filled in.
  • One change per message. It’s easier to review and undo than stacking five edits in one prompt.
  • Give real content, not placeholders. “Add a button labeled ‘Start free’ linking to temway.com/signup” beats “add a button.”
  • You can stop mid-response. If the assistant heads the wrong way, hit stop and redirect — no need to wait.
  • Keys are workspace-scoped. Anyone in the workspace can use a key once it’s added; there’s no per-seat AI billing.

Next steps