I'm Ashton, a solo developer, and I’ve built Dolphy AI—an offline, privacy-first AI assistant for Windows.
As AI becomes more powerful and pervasive, I believe it’s time to rethink how we protect our data. Dolphy AI runs entirely on-device, with no cloud calls or tracking. It’s designed for users who want the benefits of AI without sacrificing privacy.
Under the hood, Dolphy uses ONNX Runtime for fast, lightweight inference, and supports GPU/iGPU acceleration to keep things snappy. The models are tuned for GPT-4–like output while staying small enough to run on consumer hardware.
I'm Ashton, a solo developer, and I’ve built Dolphy AI—an offline, privacy-first AI assistant for Windows. As AI becomes more powerful and pervasive, I believe it’s time to rethink how we protect our data. Dolphy AI runs entirely on-device, with no cloud calls or tracking. It’s designed for users who want the benefits of AI without sacrificing privacy.
Under the hood, Dolphy uses ONNX Runtime for fast, lightweight inference, and supports GPU/iGPU acceleration to keep things snappy. The models are tuned for GPT-4–like output while staying small enough to run on consumer hardware.
I’d love your feedback—especially on performance, UX, and how we can make it better. You can check it out here: [https://apps.microsoft.com/detail/9n7094pzch4q?hl=en-GB&gl=A...]
Thanks for reading!