HRfunc

To improve human representation in science, Denny leverages linear algebra and deconvolution to estimate neural activity and hemodynamic response functions (HRFs) directly in functional near infrared spectroscopy (fNIRS) data. Leveraging a hybrid tree-hash table data structure, the HRfunc tool allows Neuroscientists to communicate observed hemodynamics, increase neural activity estimation activation, and improve represationation of people from diverse backgrounds in science. Check out the HRfunc website for more information!

AvAI™

AvAI is a artificial intelligence taught to understand snowpack characteristics and apply features learned to avalanche risk prediction. The model is undergoing training currently and expected to be released alongside a paper validating it's use in 2026. Check out the Rocky Mountain Digerati website for updates on it's progress!

Rocky Mountain Snowpack Dataset

The Rocky Mountain Snowpack dataset is a collection of snowpack images methodically collected to preserve each samples location in time & space. The datasets ultimate goal is to allow the wider articial intelligence and machine learning community to be apply their skills to snow science. The ultimate goal of this dataset is enable an AI to be trained to predict avalanche risk at higher accuracy than current avalanche risk prediction systems (~86%).

Maxim

Maxim is an agentic robotics framework that gives robots something resembling a mind. Inspired by mammalian neuroscience, it implements computational models of the hippocampus, entorhinal cortex, nucleus accumbens, and attention networks to create embodied AI agents that can perceive, remember, decide, and learn. Built for the Reachy Mini humanoid robot, Maxim processes real visual input, real motor commands, and real human interactions. Read the full overview to learn more.

synchronAI

synchronAI is a multimodal machine learning system for detecting interpersonal synchrony between two individuals. It fuses video, audio, and fNIRS (functional near-infrared spectroscopy) data to predict when two people's physiological and behavioral patterns align during interaction. By combining visual behavioral cues, vocal characteristics, and real-time brain activity measurements, synchronAI enables researchers in psychology and neuroscience to quantify a phenomenon that has historically been difficult to measure objectively. Check out the synchronAI repository on GitHub.

BOLDnet

A fMRI AI training framework for seemlessly feature engineering blood oxygenation level dependent (BOLD) data on either classification or regression problems. A foundational model examining emotional valence is undergoing training and expected to be released in the near future. Check out the BOLDnet repository on GitHub.