FaultLens — AI-Powered Resilience Engineering

FaultLens is an AI-assisted Chaos Engineering and Resilience Intelligence platform designed to help developers understand how distributed systems behave when failures occur. Modern applications are composed of many interconnected services, databases, queues, gateways, and dependencies. A failure in one component can propagate through the architecture in ways that are difficult to predict before production. Traditional monitoring often tells engineers what failed after the incident; FaultLens focuses on understanding how failures propagate and what engineers should investigate next. Users import a system architecture into FaultLens, which validates the topology and creates a simulation-ready Digital Twin represented as an interactive dependency graph. Engineers can select components and execute safe, simulated ChaosLab experiments such as service failures, latency spikes, and resource-related disruptions without touching real production infrastructure. FaultLens models failure propagation through actual system dependencies and produces experiment-specific metrics including affected nodes, blast radius, recovery behavior, metric impact, risk level, and a resilience score. Experiment parameters have real consequences in the simulation: for example, running the same failure for 10, 30, and 60 seconds can produce progressively different propagation and recovery outcomes. The platform then builds a structured FaultLensContext containing the system architecture, experiment configuration, propagation path, recovery data, resilience analysis, history, and recommendations. Through a real Model Context Protocol (MCP) integration, IBM Bob can access this context and FaultLens tools directly. Bob can inspect the latest experiment, reason about why the failure propagated, interpret the resilience result, identify critical recovery problems, and recommend the most useful next chaos experiment. This creates a continuous workflow: Architecture → Digital Twin → Chaos Experiment → Failure Propagation → Metrics → Resilience → MCP → IBM Bob → Reasoning → Recommendation FaultLens combines deterministic engineering evidence with AI reasoning: FaultLens generates the evidence; IBM Bob reasons over the evidence.