UAlberta Computing Science Students Win G7 GovAI Challenge

Akemi Izuko (left) and Steven Tang (right), MSc students with the Department of Computing Science. Photo supplied.

A team of Computing Science graduate students from the University of Alberta has won the Under-29 category at the international G7 GovAI Challenge. Akemi Izuko, Steven Tang, and Oliver Chen were recognized for their project "ReadyFormAI", an innovative tool designed to make digital government services accessible to a wider demographic.

The G7 GovAI Challenge asked participants to develop practical AI solutions to improve government communication for clients with a wide variety of needs and abilities. Competing against teams from across the G7 and the European Union, the UAlberta team focused on a major barrier for visually impaired citizens: navigating complex government PDF forms.

ReadyForm AI transforms standard, dense PDF forms into an interactive, voice-driven conversation. Built as a web extension, the tool intercepts secure government PDFs and opens them in a high-contrast interface. An AI voice assistant then guides the user through the document, offering multi-language support that covers all official G7 languages. The system navigates complex tables and auto-scrolls to the relevant fields as the user speaks, instantly translating and formatting spoken words into standardized PDF inputs.

To ensure the strict privacy standards required for handling personally identifiable information, the team engineered the system to be fully isolated. They built a self-hosted pipeline using Pipecat, NVIDIA's Parakeet model for highly accurate open-weight Speech-to-Text, and Qwen3-TTS for Text-to-Speech. This architecture allows the entire system to run securely on sovereign government infrastructure.

By removing the visual friction of traditional online forms, ReadyForm AI empowers citizens to access essential public services independently while freeing up government support workers.

Watch this video to learn more about ReadyForm AI!