Full-Stack / EnterpriseExhibited

MULTI-TENANT ADMIN PORTAL

AI-Augmented Enterprise Administration Platform

Spring Boot · Groq AI · Multi-Tenant · Flyway Migrations

The Problem Statement

Enterprise SaaS products serving multiple organisations face a recurring challenge: each tenant needs isolated data, independent user management, and role-based access — but building this from scratch for each product is cost-prohibitive. This portal (Tool-77, built for CampusPe at Aditya Birla Group) solves the multi-tenancy problem with a Spring Boot backend enforcing tenant isolation, while a Flask AI sidecar powered by Groq LLM adds intelligent content summarisation and sentiment analysis as a first-class feature. It positions itself as the operational intelligence layer between a raw database and a product team.

The Architecture Layout

The platform is a three-tier polyglot architecture. The Spring Boot backend exposes REST endpoints under JPA entity models, with Spring Security enforcing tenant-scoped access control and Redis caching for session and rate-limiting. A Flyway-managed PostgreSQL database stores tenant, user, and application data through versioned migration scripts. A Flask Python AI sidecar connects to the Groq API, exposing /describe and /recommend endpoints that process user-submitted text and return structured analysis (summary, key insights, sentiment) and actionable recommendations. The React frontend (Vite + TypeScript) communicates with both services via Axios, renders data through Recharts dashboards, and enforces role-gated navigation via React Router.

Architecture Design Diagram

MULTI-TENANT ADMIN PORTAL System Architecture

System Process Flow

User Actions──>Interface State──>Core Processing Logic
Database Persistent Layer<── Response Loop ──>Dynamic UI Updates