Livia_Zaharia

Project owner

GlucoseDAO

GlucoseDAO is a decentralized organization that enables diabetics and others with continuous glucose monitors (CGMs) to get accurate glucose predictions at least one hour in advance. For millions of diabetic people, such predictions are crucial to planning their everyday lives (when to inject, act, or do sports). It is also helpful for healthy people as it allows them to optimize their diet and exercise routines based on glucose patterns. What we are developing: -Extension of GlucoBench benchmark that will also measure human performance -Open-source ML model to predict glucose and related health outcomes -ML service and app that everybody can use easily

Ladybriggy2025

Project owner

NigerGuard:AI Climate Resilience-Vulnerable States

NigerGuard addresses critical climate vulnerabilities in Nigeria's most affected states through AI-powered prediction and community engagement. The project initially targets six highly vulnerable states: Borno, Yobe, Adamawa, Kebbi, Niger, and Lagos, where climate change impacts are most severe and documented. Northeastern Nigeria faces severe climate challenges: Borno has 78% climate-vulnerable population, 25% rainfall reduction; Yobe shows 82% vulnerability, massive Lake Chad shrinkage; Adamawa reports extensive flooding. Northwest's Kebbi faces major flood damage, while Niger State shows widespread community impact. Lagos faces coastal erosion and rising seas, risking ₦78.3B losses yearly

Luke Mahoney (MLabs)

Project owner

NEURAL SEARCH

Neural Search is a new way to incorporate Monte Carlo Tree Search into neural networks. Unlike similar techniques, Neural Search builds the exploration and optimization of the search into the layers of the network. The search itself is learnable, and provides us with a means to develop guided search algorithms trained on data. We have demonstrated that the algorithm can learn robust solutions to search problems. We propose further development of the approach to large scale problems, and to demonstrate the effectiveness of the algorithm by training it to play Go, a standard achieved AlphaGo.

Total for the project is $150K over 11 months, and it will be resourced by 3 senior MLabs members.