Software engineering courses increasingly rely on data from code hosting platforms to provide feedback about collaboration and code review behaviour. Hephaestus supports this idea by aggregating pull requests, issues, and reviews into gamified analytics for project courses. The existing system, however, is bound to a single statically configured GitHub organization, depends on a long-lived personal access token, and exposes no explicit concept of workspaces. These limitations hinder operation across multiple courses and make it difficult to reuse the platform beyond one deployment.This thesis extends Hephaestus into a multi-workspace platform with secure GitHub App based integration and explicit workspace governance. The work introduces a workspace abstraction with isolated configuration and data, separates external organizations from internal workspaces, and moves configuration such as tracked repositories and notification schedules from deployment settings into workspace-level state. The resulting architecture supports several independent workspaces within a single deployment while keeping infrastructure shared and data separated.The research follows a requirements-driven methodology. It derives functional requirements, quality attributes, and constraints for multi-workspace operation, designs an architecture around workspace context resolution and an installation-centric GitHub integration, and adapts the persistent data model accordingly. The implementation migrates the existing system to this architecture and introduces access control aligned with workspace memberships. An evaluation against the stated objectives shows that the new version can operate multiple workspaces in parallel and configure them through application interfaces rather than deployment changes.