Project HYDRA - Heuristic policY Driven Recovery Agent
Year Long Project
CompSoc
Abstract
Project HYDRA: Live Stock Monitoring is a distributed application designed to provide real-time financial market insights through a React-based dashboard and a Python FastAPI backend. The system fetches live stock data, computes key financial indicators, and ensures reliable delivery using Kubernetes-based deployment with robust routing and failover mechanisms. It incorporates a comprehensive observability setup with custom CPU and memory metrics, Prometheus monitoring, and Grafana visualization. Additionally, the project features a reinforcement learning agent trained using Proximal Policy Optimization (PPO) to enable intelligent auto-scaling and self-healing, allowing the system to maintain stability and optimize resource usage under dynamic workloads and failure conditions.
Abstract
Project HYDRA: Live Stock Monitoring is a distributed application designed to provide real-time financial market insights through a React-based dashboard and a Python FastAPI backend. The system fetches live stock data, computes key financial indicators, and ensures reliable delivery using Kubernetes-based deployment with robust routing and failover mechanisms. It incorporates a comprehensive observability setup with custom CPU and memory metrics, Prometheus monitoring, and Grafana visualization. Additionally, the project features a reinforcement learning agent trained using Proximal Policy Optimization (PPO) to enable intelligent auto-scaling and self-healing, allowing the system to maintain stability and optimize resource usage under dynamic workloads and failure conditions.
Report Information
Team Members
Team Members
Report Details
Created: April 21, 2026, 12:40 p.m.
Approved by: Sanjay S Bhat [CompSoc]
Approval date: None
Report Details
Created: April 21, 2026, 12:40 p.m.
Approved by: Sanjay S Bhat [CompSoc]
Approval date: None