Local LLM optimisation and feedback integration

At Eindhoven Engine, we have developed an AI tool called Mijke the Matchbot to assist individuals in finding the right local services. Hosted on WhatsApp and powered by a large language model, Mijke offers both voice and text interactions to ensure users receive the support they need efficiently and effectively.

Within Eindhoven Engine, a team of students and experts is addressing this issue, by approaching it through different domains such as healthcare, work, financial, education, digital, and family. This team collaborates with an ecosystem that also embraces this problem and aims to create an inclusive society. >> More about inclusive society at Eindhoven Engine.

Assignment

For this project, we are seeking a motivated student with a passion for technology, innovation, and making an impact in society. Together with the team, you will work on improving fundamental skills in language, mathematics, and digital literacy, making a difference in the lives of people in the Eindhoven region and beyond.

This project involves developing and evaluating a privacy-preserving local LLM for Mijke’s matching engine, building a feedback loop, optimizing model performance, and maintaining the system for testing and live demos.

Requirements

• Python experience
• Basic ML/NLP knowledge
• Familiarity with LLMS
• Interest in ethical AI and social impact
• Ability to work independently

 

Is your interest piqued?

Ready to harness the power of technology for positive social change? Your skills and enthusiasm can make a real difference in improving basic skills and digital literacy.