SupportIQ

May-July 2026Personal Project

A full-stack, multi-tenant AI customer support platform that grounds answers in each company’s knowledge.

Support teams need answers grounded in their own knowledge.

Built across the stack

I designed and developed the product with Java, Spring Boot, PostgreSQL, and a React/Next.js TypeScript frontend, exposing secured REST APIs for a multi-tenant SaaS architecture.

Retrieval before generation

The RAG pipeline uses OpenAI embeddings, vector search, and function calling to retrieve relevant company knowledge before generating an answer. JWT authentication and role-based access control isolate tenants and permissions.

Production discipline

JUnit and Mockito cover the backend’s core behavior, while Docker and GitHub Actions provide a repeatable path from commit to deployment.