How hawk-digital-environments/hawki works
Project type and purpose
• This is a full-stack, monolithic web application built on the Laravel framework.
• Purpose: HAWKI2 is a data-protection compliant AI-assisted group-chat and collaboration platform for universities, with modular AI-model integration (OpenAI, Google, GWDG, Ollama, etc.), export/summarization, and peer-to-peer rooms.
File structure
• Top-level “app/” directory holds the Laravel MVC backend:
• Http/Controllers → request handling (API, web)
• Models → Eloquent ORM objects (User, Room, Message, AiConv, etc.)
• Services → domain logic (AI providers, Auth integrations, UsageAnalyzer)
• Jobs & Events → background processing and broadcasting
• “resources/views” + “resources/js/css” + “public/…” = frontend presentation layer (Blade templates, vanilla JS/CSS, Vite build).
• “routes/” defines API/web/console endpoints.
• “config/” and “bootstrap/” hold environment config and framework initialization.
• “database/” for migrations and seeding; default DB is relational (MySQL/Postgres).
• Docker-related folders (“docker/”, “_docker_production/”) contain NGINX, PHP-FPM, Node build, and deployment orchestration.
• Included CLI scaffolding under “bin/_env” for environment management (developer tooling, not part of runtime architecture).
README for additional insights
• Emphasis on group chat rooms, integrated AI agents, summarization/export, and modular AI backends.
Elements
• Component labels: “Browser UI”, “NGINX Reverse Proxy”, “Laravel App (PHP-FPM)”, “Database”, “Storage”, “Queue Worker”, “AI Provider X”, etc.
• Blue rectangles for containers/services
• Orange clouds for external services
Detail—capture all runtime pieces
• Separate containers for web/proxy, app, worker, DB, cache.
• Directories in the codebase (app/Services, app/Http/Controllers) next to each backend component to tie code to runtime.