Improving the AI Assistant

Join us as we develop a safe and private AI assistant. We will begin by researching the required software stack.

This paper will discuss the software stack, the best practices around security, and ethical considerations. It shows the range of work that has been and will be done in this project; there is something for every software developer and data scientist to work on.

This is the current state of implementation and consideration for the next step, so it will be updated as the project progresses.

The languages used in the software stack

First, let's take a high-level view of the software stack. The currently used languages are Python, C++, Rust, Mojo, PHP, JavaScript, Go, Dart, Julia, Scala, and Elixir. It is not an attempt to tick as many languages as possible but just the necessary dependencies from used frameworks and libraries. The choice of software components and how custom components are written are governed according to the best-known method of solving a specific function in the AI Assistant solution. The main objective is to ensure security, stability, and optimal performance. The analogy to using the best tool for the job is to choose the language with the best fit for the function. Small teams with limited skills often rely on familiar solutions, as if every problem is a nail for their hammer.

However, it is essential to consider the chosen languages' and components' scalability and maintainability. Additionally…