Facultad de Comunicaciones y Filología, Universidad de Antioquia operates multiple administrative and academic processes that evolved over years of incremental development. Rather than a single application, the faculty relies on an interconnected ecosystem of services supporting administrative and academic workflows, institutional reporting and identity management.

The engineering challenge is not building one greenfield product: it is coordinating legacy modernization, shared authentication, consistent data workflows and new interfaces while keeping critical processes available to staff and students.

Several production systems were built on older Laravel and Symfony codebases with long-lived business rules, custom database schemas and tightly coupled UI flows. These applications remain essential for day-to-day faculty operations and cannot be replaced overnight.

Work on the legacy layer focuses on stabilizing behavior, improving maintainability, documenting critical paths and preparing modules for gradual extraction into APIs or modern frontends, without disrupting institutional users.

Newer components introduce clearer separation between frontend and backend, typed interfaces and containerized development environments. React and Next.js applications consume Laravel APIs, while Python services handle specialized reporting workloads.

The modern layer is designed to grow alongside legacy systems, sharing authentication, reusing validated data models where possible and exposing capabilities through documented APIs rather than monolithic screens.

Identity management is centralized through Keycloak with OIDC, allowing multiple applications to share a single sign-on experience without duplicating user stores or session logic in every codebase.

Integration work spans legacy PHP applications, modern JavaScript frontends and service-to-service flows, with careful attention to token lifecycles, role mapping and federated access patterns appropriate for an institutional environment.

Faculty operations generate structured and semi-structured data that must be exported, reviewed and delivered to administrative teams. Python reporting services handle Excel generation, data aggregation and asynchronous processing without blocking interactive applications.

Reporting pipelines emphasize traceable transformations, validation scripts and repeatable outputs, treating reports as production artifacts, not ad hoc scripts.

The ecosystem is modernized incrementally: identify bounded modules, expose them through APIs, replace UI surfaces one at a time and validate each step before moving dependencies. This reduces risk compared to a full rewrite while still moving toward maintainable architecture.

Documentation, validation scripts and Docker environments support teams working across PHP, Python and TypeScript, making onboarding and regression checks part of the standard workflow.