2025-09-01

Multi-Tenant Helpdesk AI SaaS (RAG)

Stack: Django / RAG / OpenAI / PostgreSQL

Multi-tenant AI helpdesk with PDF ingestion, vector retrieval (FAISS), and OpenAI-grounded answers.

Highlights

• ✓Designed and developed a multi-tenant AI-powered chatbot SaaS with strict tenant-level data isolation (schema-based tenancy).
• ✓Implemented a RAG pipeline to answer user queries from tenant-specific PDF documents.
• ✓Built PDF ingestion: text extraction, chunking, and vector embedding generation.
• ✓Stored embeddings in FAISS for semantic retrieval; integrated OpenAI for grounded responses.
• ✓Developed REST APIs for chat sessions, message history, and PDF management using Django REST Framework.
• ✓Implemented role-based access control for tenant users and data.
• ✓Containerized the application with Docker for consistent deployment.