| I spent the last year building Eurora to not have to explain the context of what I am doing, every time I wanted to ask a question. Eurora is a cross-platform application that creates a custom network layer between itself and every single browser in existence and runs on Linux, macOS and Windows. This allows the AI assistant to interact with the browser and see the whole website. As well as run mcp-like commands against the website you’re currently on. I also built a number of custom adapters. For example, asking a question about a video on YouTube allows the AI to retrieve the transcript of the video you’re watching, as well as the frames and other information like the current timestamp. The current timestamp also allows the AI to then understand the line that the person in the video just said. We also have adapters for Twitter and Google Docs right now to be able to retrieve structured data. Eurora works on every single website already by using standard calls and strategies. Eurora is built to run on a dedicated machine and has a separate server component for things like cron-jobs, indexing as well as all kinds of various processing in the future. The application is local-first and can be run on local hardware without ever touching external servers. If you want to use certain models that are too big to run locally, then you can connect to our Sovereign European Cloud. We specifically designed the server code in such a way that you can see exactly how your data is accessed (spoiler: it’s not). The goal here is to provide a fully secure and private cloud LLM environment that we can prove is fully secure. You can find out more about Eurora below: Video demo - https://youtu.be/fj8cmNu_c5Y Github - https://github.com/eurora-labs/eurora You can download our app for every platform and every browser below. You get 500,000 cloud tokens for free when creating an account, it would be immensely helpful if you could tell me what you think of Eurora. Website - https://www.eurora-labs.com Download link - https://www.eurora-labs.com/download [link] [comments] |