Your phone finishes your sentences. Your inbox writes its own replies.
What should this do for AAC, without giving up your privacy?
We're open-sourcing an AAC app and inviting AAC users, SLPs, parents, and educators to trial it and help us figure out the answer. No locked roadmap. No selling your data. Just an honest invitation to shape what comes next.
Open questions we're testing with you.
Large Language Models. Machine Learning. Natural Language Processing. The same tech behind autocomplete, smart reply, and voice assistants. Here's what we want to bring to AAC, with you.
Privacy first. Everything below is off by default.
Nothing collects, predicts, or learns until you (or an authorized supporter, for a minor or supported user) turn it on. The capabilities are real enough that you'll want some of them. The choice stays yours.
- Predictions that learn how you communicate. Faster sentences in your own style, on your own device.
- Session notes that draft themselves. Spend your hours with students, not with paperwork.
- Insight for SPED directors and clinical leads. Aggregate signal across a district, never surveillance of any one student.
- Modeling support that fits your family. Prompts and visuals shaped to your routine, not a generic curriculum.
Your communication belongs to you.
- We do not sell your data. Not now. Not ever.
- No training on your communication without an explicit yes. Opt in, opt out, on your terms.
- Optional anonymized data helps make predictions better for the whole AAC community. You decide.
- Open source. The roadmap and the code live on GitHub.