01
Availability
Date-based inventory connects room types, concrete rooms, and stay ranges.
The Sailing Bay · 2025
Hotel
Management.
A completed Django academic capstone: a server-rendered monolith with SQLite for booking, inventory, cancellation, and administration, alongside a retrieval-augmented knowledge assistant.
DjangoPythonSQLiteServer-renderedRAGFAISS
View source 
2025.10 → 12
Development window
6
Focused tests
SQLite
Database
Completed
Capstone status
Verified scope
A booking system with clear boundaries.
02
Reservations
A booking lifecycle covers creation, overlap rejection, and cancellation.
03
Inventory
Booking signals decrement inventory and cancellation restores affected rows.
04
Account flows
Authenticated users can access profile and booking history workflows.
05
Staff operations
Dashboard summaries and check-in or check-out actions support staff access.
06
Hotel assistant
A retrieval-augmented path prepares hotel knowledge for an LLM response.
Architecture
MVT at the core, RAG at a separate boundary.
Django MVT monolithURL configuration → views and ORM → server-rendered templates
↓ Domain workflows
Booking domainRooms · inventory · reservations · cancellations
SQLiteUsers · rooms · dates · booking records
Selected ORM loadingConcrete select_related usage in targeted queries
↓ Knowledge retrieval path
Knowledge base → HuggingFace embeddings → FAISS → retrieved context → LLM responseGroq cloud path or optional local LlamaCpp / Phi-4 path
The two provider paths are separate implementation options. There is no automatic fallback, asynchronous inference, or verified production AI runtime.
Engineering decisions
Consistency before impressive numbers.
01
transaction.atomic and select_for_update wrap the documented booking conflict check.
02
A booking signal decrements inventory for the stay range, while cancellation restores affected inventory rows.
03
Selected relationship loading is presented as a concrete ORM technique, not a claim of global query optimization.
Repository timeline
Built across three documented milestones.
2025.10.28First repository commit
2025.11.19Final version commit
2025.12.04Groq integration added
Test evidence
Six tests around core behavior.
The documented baseline covers booking creation and inventory decrement, overlap rejection, cancellation restoration, user ownership, missing chatbot input, and sanitized chatbot failure behavior.
python manage.py checkpython manage.py testThe tests mock the chatbot boundary and do not call an external AI provider. This is a meaningful baseline, not comprehensive coverage.
Visual evidence boundary
No proof is manufactured from what is missing.
The repository does not track screenshots of booking/search, room detail, admin dashboard, or chatbot conversations. Room and service imagery is content material, not UI evidence.
Need recapture
Booking/search, room detail, admin, chatbot
Ready as support
Room and service gallery content
Do not use
Stale ERD and removed media
Ownership
Student Project Owner / Maintainer
The student project owner directed requirements, capstone scope, feature decisions, testing, and academic delivery. AI assistance supported implementation, refactoring, architecture proposals, debugging, and testing assistance.