resilience-analysis

Fast and scalable fault resilience analysis for deep neural networks

a Go library providing resilience patterns to make your services more robust and fault-tolerant. Built to endure failures with battle-tested patterns like Circuit Breaker and Exponential Backoff Retry.

Web service for designing a resilient LEO satellite constellation and measuring end-to-end reachability from Arctic ground terminals to a gateway. Vectorised orbital geometry, graph-based availability analysis and a configuration optimiser. 4th of 55 at КосмоХакатон 2026.

EPITA group project, anomaly detection

A C++ implementation of an analysis tool for water supply networks, enabling maximum flow computation, deficit detection, and failure simulation in an interactive environment, developed for the Algorithm Design course at FEUP.

Applying Graph Neural Networks (GNNs) to predict congestion and analyze resilience in complex logistics hubs. Features dynamic graph modeling, spatio-temporal forecasting, MLOps deployment on GCP Vertex AI, and API integration. Demonstrates end-to-end ML system development.