How Token Terminators Made AI Greener with Smarter Prompts
_web.jpg)
_web.jpg)
_web.jpg)
The team discovered that large language models like ChatGPT consume nearly ten times more energy than traditional search engines. With AI becoming a part of everyday life, they saw a clear opportunity to reduce its environmental impact by focusing on one of the most basic interactions: the prompt.
_web.jpg)
_web.jpg)
To address this, they developed a full-stack application that intelligently compresses user input before it reaches the model. The backend identifies and retains only the essential elements of the message, while the front end allows users to test the results in real time. This streamlined process reduces energy usage, improves performance, and lowers operational costs — all without compromising the quality of the output.
The solution impressed the judges and secured the team a well-earned win.
As we look ahead to the BCG Platinion Hackathon 2025, we’re excited to see what bold ideas this year’s participants will bring. The theme, Future of Education & Learning, invites teams to explore how technology can reshape the way we teach, learn, and grow.
Applications are now open. Apply here.