Borkov

Borkov is an agent-powered simulation and analytics engine for DeFi lending protocols, designed to optimize fixed-term lending markets using Markov processes and machine learning. Our project models user behavior archetypes—such as Diamond Hands, Paper Hands, and Degen Traders—by simulating loan repayment probabilities over time and visualizing the evolution of individual and aggregate lending multipliers. At its core, Borkov builds a Markov state machine to track transitions between loan states: Reset (new), Building (growing trust), and Established (proven reliability). By varying transition probabilities, Borkov reveals systemic risk, reward distributions, and agent behavior under diverse market conditions. The platform provides interactive dashboards: visualizing multiplier distributions, risk profiles, state evolution, and journey trajectories across thousands of simulated lending sessions. Our engine calculates steady-state probabilities, risk spreads, and credit trajectories, offering protocol designers unique, data-driven insight into sustainable lending models, risk segmentation, and incentive mechanism design. Ultimately, Borkov aims to make DeFi lending safer and more predictive through statistical agent simulation and advanced analytics.