How ScrollEngine's AI features use your data.

⟦LEGAL⟧ Working draft for counsel review. Not in effect. This statement was prepared for counsel under gate G5 (00 §2.2) and the AI data-use decision (FD-11). It will be replaced by the signed text before it is published.

Open points for counsel and engineering: the training sentence in section 5, which is published only if it holds for every provider path; the platform-default provider; the exact records each AI tool can read.

1. Which features this statement covers

• Ask Engine ◐ Early access, our AI operations assistant. It answers questions about live orders, drivers, routes, exceptions and cash on delivery, and links each answer to the records behind it. It runs in the ScrollEngine web dashboard and is in English for now. Ask Engine inside Shopify admin ◌ Vision is not available.
• AI steps in workflows ◐ Early access, which summarize, classify or investigate inside workflows that your team sets up.
• Bring your own model ● Live: you choose the model provider these features use.

2. What AI features can read

AI features read only what your account, and each user's permissions, allow. Each answer is limited to what the person asking is allowed to see. Usage is limited by a daily cap.

3. What AI features can't do

Today our AI can read, not write. No AI feature changes an order, a route or a message. Workflows act only through the steps your team configures, and every rule, route and workflow works with AI turned off.

These aren't AI, and we don't label them that way: route optimization, automated dispatch, live traffic-aware ETAs, at-risk alerts, location priority rules and driver scorecards. They run on rules, optimization and live data.

4. Which model providers are used

You choose. Bring your own key for OpenAI, Anthropic, Google Gemini or OpenRouter, or use the platform default, OpenRouter (as of Oct 2026). When you use an AI feature, the question and the records needed to answer it are sent to that provider.

• With the platform default, the provider processes the data as our sub-processor, listed on our sub-processors page and covered by our data processing addendum.
• With your own model key, your agreement with that provider also applies.

5. Training and benchmarks

We don't use your data to train third-party models. Aggregated, anonymized benchmarks use your data only with your consent.

6. Turning AI off

AI features are optional. If your policy doesn't allow operational data to go to an outside model provider, leave them off: slots, routes, drivers, tracking and workflows all keep working.

7. Changes to this statement

When we change how AI features use data, we'll update this statement with a new effective date and version number, and tell account owners about material changes before they take effect.

8. Contact

Questions about AI and your data: support@scrollengine.com (as of Oct 2026), or contact us. The rest of our security and privacy controls are on the trust center.