Zuno AI

Inspiration

We are studying data science, so we spend a lot of time writing Python code and managing Git repositories. But we also have creative hobbies. Jumping between coding screens and creative workspaces felt really messy. We wanted one voice-controlled assistant to handle our computer so we could stay focused while working on all these different things.

What it does

Zuno AI is a voice assistant that actually controls your Windows computer. It does not just chat with you in a browser. It can open apps, type notes, and even make WhatsApp calls for free by clicking the screen for you. It also has a "Creative Mode." In this mode, it can write stories or brainstorm ideas, and it reads them out loud using a custom voice player.

How we built it

We built the whole project in Python. The brain of the assistant uses the Google Gemini API to understand what we are saying. For the design, we used a library called CustomTkinter to make a sleek, dark-themed dashboard. To make Zuno actually control the computer, we used PyAutoGUI to click on the screen and type, and the subprocess module to run system commands directly.

Challenges we ran into

The biggest problem was WhatsApp. The normal code couldn't hang up calls properly because of how the app runs. We had to write a strict, direct system command to kill the process completely. Another big issue was the voice engine. When Zuno generated long stories with weird characters, the audio would crash. We fixed this by having Zuno save the text to a temporary file first, and then asking Windows to safely read that file in the background.

Accomplishments that we're proud of

We are really proud that Zuno can automate WhatsApp calls and messages for free without needing a paid API. We are also really happy with how the app looks. Upgrading the interface to a modern design without breaking the voice or the background tasks was a huge win for us

What we learned

We learned that just using Python libraries isn't always enough for desktop automation. Sometimes, you have to talk directly to the operating system to get things done. It actually reminded us of our classes on digital logic and low-level programming—sometimes you just need to get closer to the hardware and run strict system commands to make things work reliably.

What's next for Zuno AI

Right now, Zuno clicks on exact screen coordinates. If a window moves, it misses the click. Next, we want to give it "Screen Vision" so it can actually see where buttons are. We also want to add a safe space for it to run its own Python scripts. Finally, we plan to add voice-controlled Git commands, so we can just tell it to push code directly from the present working directory.