braai (Bogus/Real Adversarial AI) performs deep-learning real/bogus classification for the Zwicky Transient Facility (ZTF), separating genuine astrophysical events and objects from false positive detections. It uses a convolutional neural network to enable efficient automated detection of flux transients, recurring flux-variable sources, and moving objects in large-scale astronomical survey data. In production, it achieves low false negative and false positive rates.

AMPEL provides an analysis framework for high-throughput surveys and is suited for streamed data. The package combines the functionality of an alert broker with a generic framework capable of hosting user-contributed code; it encourages provenance and keeps track of the varying information states that a transient displays. The latter concept includes information gathered over time and data policies such as access or calibration levels.

The Bayesian-based Gaussian Process model AsteroGaP (Asteroid Gaussian Processes) fits sparsely-sampled asteroid light curves. By utilizing a more flexible Gaussian Process framework for modeling asteroid light curves, it is able to represent light curves in a periodic but non-sinusoidal manner.

Tails identifies and localizes comets in image data from the Zwicky Transient Facility (ZTF), a robotic optical sky survey, using deep-learning with a custom EfficientDet-based architecture. It detects comets in single images in near real time, rather than requiring multiple epochs as in traditional methods. In production, Tails achieves 99% recall, a false positive rate below 0.01%, and 1–2 pixel root mean square error in the predicted position.