# cataverse.ai

cataverse.ai is an autonomous research platform for catalyst discovery, built to shift development from trial-and-error to predictive design. The scientific goal is to understand how operating conditions reshape oxide-supported transition-metal catalysts in real time, then use that understanding to engineer materials with targeted structure and function. By treating catalyst evolution as a controllable process rather than a post hoc observation, the platform supports faster iteration and more reproducible materials development.

The core approach combines in situ instrumentation, closed-loop experiment execution, and model-informed decision making. Vibrational spectroscopy and complementary measurements are used to track adsorbate-driven atom mobility, nucleation, and metastable phase evolution, while the resulting kinetics and state variables feed an evolving digital representation of the system. This enables adaptive experiment selection, multivariable phase mapping, and operational strategies intended to mitigate undesired structural drift (for example, sintering) during both synthesis and use.

This repository currently contains two coordinated subpackages: orchestration/, which handles experiment orchestration and instrument-control workflows, and ir-spectro-node/, which serves as the spectroscopy node in the distributed platform. Each subpackage has its own README with implementation and usage details.

At present, the instrument is operating in batch experimental mode only. The active hardware configuration is centered on a vacuum manifold, which supports the current closed-loop batch workflows for synthesis, treatment, and characterization. A dedicated flow manifold is under active integration and is expected online soon, enabling continuous-flow experimental modes alongside the existing batch capability.